Upload
dinhtruc
View
217
Download
4
Embed Size (px)
Citation preview
Hindawi Publishing CorporationISRN BiomathematicsVolume 2013 Article ID 403549 7 pageshttpdxdoiorg1011552013403549
Research ArticleOptimal Control of an SIR Model with Delay in State andControl Variables
Mohamed Elhia Mostafa Rachik and Elhabib Benlahmar
Laboratory of Analysis Modeling and Simulation Department of Mathematics and Computer Science Faculty of Sciences Ben MrsquoSikHassan II University Mohammedia BP 7955 Sidi Othman Casablanca Morocco
Correspondence should be addressed to Mohamed Elhia elhia mohamedyahoofr
Received 30 May 2013 Accepted 11 July 2013
Academic Editors H Ishikawa and M A Panteleev
Copyright copy 2013 Mohamed Elhia et alThis is an open access article distributed under the Creative CommonsAttribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited
We will investigate the optimal control strategy of an SIR epidemic model with time delay in state and control variables We usea vaccination program to minimize the number of susceptible and infected individuals and to maximize the number of recoveredindividuals Existence for the optimal control is established Pontryaginrsquos maximum principle is used to characterize thisoptimal control and the optimality system is solved by a discretization method based on the forward and backward differenceapproximations The numerical simulation is carried out using data regarding the course of influenza A (H1N1) in Morocco Theobtained results confirm the performance of the optimization strategy
1 Introduction
For a long time infectious diseases have caused several epi-demics leaving behind them not only millions of dead andinfected individuals but also severe socioeconomic conse-quences Nowadays mathematical modeling of infectiousdiseases is one of the most important research areas Indeedmathematical epidemiology has contributed to a betterunderstanding of the dynamical behavior of infectious dis-eases its impacts and possible future predictions about itsspreading Mathematical models are used in comparingplanning implementing evaluating and optimizing variousdetection prevention therapy and control programs Manyinfluential results related to the development and analysisof epidemiological models have been established and can befound in many articles and books (see eg [1ndash3])
Epidemiologicalmodels often take the formof a systemofnonlinear ordinary and differential equations without timedelay However for various biological reasons the realdynamic behavior of an epidemic depends not only on its cur-rent state but also on its past history Thus to reflect the realbehavior of some diseases many researchers have proposedand analyzedmore realisticmodels including delays tomodel
differentmechanisms in the dynamics of epidemics like latentperiod temporary immunity and length of infection (see eg[4ndash8] and the references therein)
To the best of our knowledge including time delay inboth state and control variables in the context of an epidemiccontrolledmodel has not been studiedThere have been someworks (like [9 10]) which study an optimal control problemwith time delay but only in the state variable In this paper wewill investigate the effect of a vaccination program in the caseof an SIR (susceptible-infected-recovered) epidemic modelwith time delay in the control and the state variables To dothis we will consider an optimally controlled SIR epidemicmodel with time delay where the control means the percent-age of susceptible individuals being vaccinated per time unitand the time delay represents the required time so that a vac-cinated susceptible person moves from the susceptibles classto the recovered class We use optimal control approach tominimize the number of susceptible and infected individualsand to maximize the number of recovered individuals duringthe course of an epidemic
This paper is organized as follows In Section 2 we willpresent a mathematical model with time delay and a controlterm The analysis of optimization problem is presented in
2 ISRN Biomathematics
Section 3 In Section 4 we will give a numerical appropriatemethod and the corresponding simulation results Finally theconclusions are summarized in Section 5
2 Mathematical Model
We consider the SIR epidemic model with constant totalpopulation sizeThe population is divided into three disease-state compartments susceptible individuals (119878) people whocan catch the disease infectious (infective) individuals (119868)people who have the disease and can transmit the diseaserecovered individuals (119877) people who have recovered fromthe disease We assume that an individual can be infectedonly through contacts with infectious individuals and thatimmunity is permanent The transitions between differentstates are described by the following parameters
(i) Λ is the recruitment rate of susceptibles
(ii) 120573 is the effective contact rate
(iii) 119889 is the natural mortality rate
(iv) 120574 is the recovery rate
(v) 120598 is the disease induced death rate
The population dynamics is given by the following system ofordinary differential equations subject to nonnegative initialconditions
119889119878 (119905)
119889119905= Λ minus 120573119878 (119905)
119868 (119905)
119873 (119905)minus 119889119878 (119905)
119889119868 (119905)
119889119905= 120573119878 (119905)
119868 (119905)
119873 (119905)minus (120574 + 119889 + 120598) 119868 (119905)
119889119877 (119905)
119889119905= 120574119868 (119905) minus 119889119877 (119905)
(1)
where 119878(0) = 1198780 119868(0) = 119868
0119877(0) = 119877
0 and119873(119905) = 119878(119905)+119868(119905)+
119877(119905) is the total population number at time 119905The strategy of the control we adopt consists of a vaccina-
tion program our goal is to minimize the level of susceptibleand infected individuals and to maximize the recoveredindividuals Into the model (1) we include a control 119906 thatrepresents the percentage of susceptible individuals beingvaccinated per time unit In order to have a realisticmodel weneed to take into account that themovement of the vaccinatedsusceptible individuals from the class of susceptibles into therecovered class is subject to delay Thus the time delay isintroduced in the system as follows at time 119905 only a per-centage of susceptible individuals that have been vaccinated120591 time unit ago that is at time 119905 minus 120591 are removed from thesusceptible class and added to the recovered class So themathematical system with time delay in state and control
variables is given by the nonlinear retarded differentialequations
119889119878 (119905)
119889119905= Λ minus 120573119878 (119905)
119868 (119905)
119873 (119905)minus 119889119878 (119905) minus 119906 (119905 minus 120591) 119878 (119905 minus 120591)
119889119868 (119905)
119889119905= 120573119878 (119905)
119868 (119905)
119873 (119905)minus (120574 + 119889 + 120598) 119868 (119905)
119889119877 (119905)
119889119905= 120574119868 (119905) minus 119889119877 (119905) + 119906 (119905 minus 120591) 119878 (119905 minus 120591)
(2)
In addition for biological reasons we assume for 120579 isin [minus120591 0]that 119878(120579) 119868(120579) and 119877(120579) are nonnegative continuous func-tions and 119906(120579) = 0 Note that the control 119906 is assumed to beintegrable in the sense of Lebesgue bounded with 0 le 119906 le119887 lt 1 and 119887 is a given constant
To show the existence of solutions for the control system(2) we first prove that the system (2) is dissipative that is allsolutions are uniformly bounded in a proper subset Ω isin R3
+
Let (119878 119868 119877) isin R3+be any solution with nonnegative initial
conditions Adding equations of (2) we get
119889119873
119889119905= Λ minus 119889119873 minus 120598119868 lt Λ minus 119889119873 (3)
After integration using the constant variation formula wehave
119873(119905) leΛ
119889+ 119873 (0) 119890
minus119889119905 (4)
It then follows that
0 le 119873 (119905) leΛ
119889as 119905 997888rarr infin (5)
Therefore all feasible solutions of the system (2) enter into theregion
Ω = (119878 119868 119877) isin R3
+ 119873 le
Λ
119889 (6)
Then we can rewrite (2) in the following form
119889119883
119889119905= 119860119883 + 119865 (119883119883
120591) = 119866 (119883119883
120591) (7)
where
119883 (119905) = [
[
119878 (119905)
119868 (119905)
119877 (119905)
]
]
119860 = [
[
minus119889 0 0
0 minus (119889 + 120574 + 120576) 0
0 120574 minus119889
]
]
119865 (119883 (119905) 119883120591(119905)) =
[[[[[[
[
Λ minus120573119878 (119905) 119868 (119905)
119873 (119905)minus 119906120591(119905) 119878120591(119905)
120573119878 (119905) 119868 (119905)
119873 (119905)
119906120591 (119905) 119878120591 (119905)
]]]]]]
]
119906120591(119905) = 119906 (119905 minus 120591) 119883
120591(119905) = 119883 (119905 minus 120591)
(8)
ISRN Biomathematics 3
The second term on the right-hand side of (7) satisfies1003816100381610038161003816119865 (1198831 (119905) 1198831120591 (119905)) minus 119865 (1198832 (119905) 1198832120591 (119905))
1003816100381610038161003816
le 1198721
10038161003816100381610038161198831 (119905) minus 1198832 (119905)1003816100381610038161003816
+ 1198722
10038161003816100381610038161198831120591 (119905) minus 1198832120591 (119905)1003816100381610038161003816
(9)
where1198721and119872
2are some positive constants independent
of the state variables 119878(119905) 119868(119905) and 119877(119905) and10038161003816100381610038161198831 (119905) minus 1198832 (119905)
1003816100381610038161003816 =10038161003816100381610038161198781 (119905) minus 1198782 (119905)
1003816100381610038161003816
+10038161003816100381610038161198681 (119905) minus 1198682 (119905)
1003816100381610038161003816 +10038161003816100381610038161198771 (119905) minus 1198772 (119905)
1003816100381610038161003816
10038161003816100381610038161198831120591 (119905) minus 1198832120591 (119905)1003816100381610038161003816 =
10038161003816100381610038161198781120591 (119905) minus 1198782120591 (119905)1003816100381610038161003816
+10038161003816100381610038161198681120591 (119905) minus 1198682120591 (119905)
1003816100381610038161003816 +10038161003816100381610038161198771120591 (119905) minus 1198772120591 (119905)
1003816100381610038161003816
(10)
Here
119878119894120591 (119905) = 119878119894 (119905 minus 120591) 119868
119894120591 (119905) = 119868119894 (119905 minus 120591)
119877119894120591(119905) = 119877
119894(119905 minus 120591) for 119894 = 1 2
(11)
Moreover we get1003816100381610038161003816119866 (1198831 1198831120591) minus 119866 (1198832 1198832120591)
1003816100381610038161003816 le 119872 (10038161003816100381610038161198831 minus 1198832
1003816100381610038161003816 +10038161003816100381610038161198831120591 minus 1198832120591
1003816100381610038161003816)
(12)
where
119872 = max (1198721+ 119860 1198722) lt infin (13)
Thus it follows that the function119866 is uniformly Lipschitzcontinuous From the definition of the control 119906(119905) and therestriction on 119878(119905) 119868(119905) and 119877(119905) ge 0 we see that a solutionof the system (7) exists (see [11])
3 The Optimal Control Problem
Our goal is to minimize the number of susceptible andinfected individuals and to maximize the number of recov-ered individuals during the course of an epidemic Mathe-matically for a fixed terminal time 119905
119891 the problem is to min-
imize the objective functional
119869 (119906) = int
119905119891
0
1198601119878 (119905) + 119860
2119868 (119905) minus 119860
3119877 (119905) +
1198604
21199062(119905) 119889119905
(14)
where 119860119894ge 0 (for 119894 = 1 4) denote weights that balance
the size of the terms In other words we seek the optimalcontrol 119906lowast such that
119869 (119906lowast) = min 119869 (119906) 119906 isin U (15)
whereU is the set of admissible controls defined by
U = 119906 0 le 119906 le 119887 lt 1 119905 isin [0 119905119891]
119906 is Lebesgue measurable (16)
31 Existence of an Optimal Control The existence of theoptimal control can be obtained using a result by Fleming andRishel in [12]
Theorem 1 Consider the control problem with system (2)There exists an optimal control 119906lowast isin U such that
119869 (119906lowast) = min119906isinU
119869 (119906) (17)
Proof To use an existence result in [12] we must check thefollowing properties
(1) The set of controls and corresponding state variablesis nonempty
(2) The control setU is convex and closed
(3) The right-hand side of the state system is bounded bya linear function in the state and control variables
(4) The integrand of the objective functional is convex onU
(5) There exist constants 1198881 1198882gt 0 and 120588 gt 1 such that
the integrand 119871(119878 119868 119877 119906) of the objective functionalsatisfies
119871 (119878 119868 119877 119906) ge 1198882+ 1198881(|119906|2)1205882
(18)
An existence result by Lukes [13] was used to give the exis-tence of solution of system (2) with bounded coefficientswhich gives condition 1 The control set is convex and closedby definition Since the state system is linear in 119906 the rightside of (2) satisfies condition 3 using the boundedness ofthe solution The integrand in the objective functional (14) isconvex onU In addition we can easily see that there exist aconstant 120588 gt 1 and positive numbers 119888
1and 1198882satisfying
119871 (119878 119868 119877 119906) ge 1198882+ 1198881(|119906|2)1205882
(19)
32 Characterization of the Optimal Control In order toderive the necessary condition for the optimal controlPontryaginrsquos maximum principal with delay given in [14] wasusedThis principal converts (2) (14) and (15) into a problemof minimizing a Hamiltonian119867 defined by
119867 = 1198601119878 (119905) + 119860
2119868 (119905) minus 119860
3119877 (119905) +
1198604
21199062(119905) +
3
sum
119894=1
120582119894119891119894
(20)
where 119891119894is the right side of the differential equation of the 119894th
state variable
4 ISRN Biomathematics
Theorem 2 Given an optimal control 119906lowast isin U and solutions119878lowast 119868lowast and 119877lowast of the corresponding state system (2) there existadjoint functions 120582
1 1205822 and 120582
3satisfying
1= minus 119860
1+ 1205821119889 + (120582
1minus 1205822) 120573
119868lowast
119873lowast
+ 120594[0119905119891minus120591]
(119905) (120582+
1minus 120582+
3) 119906lowast
2= minus119860
2+ (1205821minus 1205822) 120573
119878lowast
119873lowast+ (119889 + 120574 + 120598) 120582
2minus 1205741205823
3= 1198603+ 1205823119889
(21)
with the transversality conditions
1205821(119905119891) = 1205822(119905119891) = 1205823(119905119891) = 0 (22)
Furthermore the optimal control 119906lowast is given by
119906lowast(119905) = min(119887max(0
(120582+
1minus 120582+
3)
1198604
120594[0119905119891minus120591]
(119905) 119878lowast)) (23)
where 120582+119894(119905) = 120582
119894(119905 + 120591) for 119894 = 1 3
Proof The adjoint equations and transversality conditionscan be obtained by using Pontryaginrsquos maximum principlewith delay in the state and control variables [14] such that
1= minus
120597119867 (119905)
120597119878 (119905)minus 120594[0119905119891minus120591]
(119905)120597119867 (119905 + 120591)
120597119878 (119905 minus 120591) 120582
1(119905119891) = 0
2= minus
120597119867
120597119868 120582
2(119905119891) = 0
3= minus
120597119867
120597119877 120582
3(119905119891) = 0
(24)
The optimal control 119906lowast can be solved from the optimalitycondition
120597119867
120597119906+ 120594[0119905119891minus120591]
(119905)120597119867 (119905 + 120591)
120597119906120591
= 0 (25)
That is
1198604119906 + 120594[0119905119891minus120591]
(119905) (120582+
3minus 120582+
1) 119878 = 0 (26)
By the bounds in U of the control it is easy to rewrite 119906lowast inthe form (23)
4 Numerical Simulation
In this section we give a numerical method to solve the opti-mality system which is a two-point boundary value problemwith separated boundary conditions at times 119905
0= 0 and 119905
119891
Let there exist a step size ℎ gt 0 and (119899119898) isin N2 with120591 = 119898ℎ and 119905
119891minus 1199050= 119899ℎ For reasons of programming we
0 30 60 90 120 150 1800
05
1
15
2
25
3
Time (days)
With controlWithout control
Susc
eptib
les
S(t)
times107
Figure 1 The function 119878 with and without control
consider 119898 knots to left of 1199050and right of 119905
119891 and we obtain
the following partition
Δ = (119905minus119898= minus120591 lt sdot sdot sdot lt 119905
minus1lt 1199050= 0 lt 119905
1lt sdot sdot sdot lt 119905
119899
= 119905119891lt 119905119899+1
lt sdot sdot sdot lt 119905119899+119898
)
(27)
Then we have 119905119894= 1199050+ 119894ℎ (minus119898 le 119894 le 119899 + 119898) Next we
define the state and adjoint variables 119878(119905) 119868(119905) 119877(119905) 1205821 1205822
1205823 and the control 119906 in terms of nodal points 119878
119894 119868119894 119877119894 1205821198941 1205821198942
120582119894
3 and 119906
119894 Now using combination of forward and backward
difference approximations we obtain the Algorithm 1The numerical simulations were carried out using data
regarding the course of the influenza A (H1N1) in MoroccoThe initial conditions and parameters of the system (2) aretaken from [15 16] while the time delay value is taken from[17]
120573 = 03095 Λ = 117417
119889 = 39139 times 10minus5 120574 = 02
120576 = 00063 120591 = 10
(28)
the initial conditions for the ordinary differential systemwere
119878 (0) = 30 times 106 119868 (0) = 30 119877 (0) = 28 (29)
and the transversality conditions for the ordinary differentialsystem were
120582119894(119905119891) = 0 (119894 = 1 3) (30)
Figure 1 indicates that the number of susceptible indi-viduals (119878) decreases more rapidly in the case with controlIt reaches 7399 times 105 at the end of the vaccination periodagainst 6732 times 106 in case without control that is a reduc-tion of 5992 times 106
ISRN Biomathematics 5
Step 1 For 119894 = minus119898 0 do 119878119894= 1198780 119868119894= 1198680 119877119894= 1198770 119906119894= 0
End forFor 119894 = 119899 119899 + 119898 do 120582119894
1= 0 120582119894
2= 0 120582119894
3= 0
End forStep 2 For 119894 = 0 119899 minus 1 do
119878119894+1=119878119894+ ℎ(Λ minus 119906
119894minus119898119878119894minus119898)
1 + ℎ(120573(119868119894119873) + 119889)
119868119894+1=
119868119894
1 + ℎ(120574 + 119889 + 120576 minus 120573(119878119894+1119873))
119877119894+1=119877119894+ ℎ(120574119868
119894+1+ 119906119894minus119898119878119894minus119898)
1 + ℎ119889
120582119899minus119894minus1
1=
120582119899minus119894
1+ ℎ(119860
1+ 120573(119868
119894+1119873)120582
119899minus119894
2+ (120582119899minus119894+119898
3minus 120582119899minus119894minus1+119898
1)120594[0119905119891minus120591]
(119905119899minus119894)119906119894)
1 + ℎ(119889 + 120573(119868119894+1119873))
120582119899minus119894minus1
2=120582119899minus119894
2+ ℎ(119860
2+ 120582119899minus119894
3120574 minus 120582119899minus119894minus1
1120573(119878119894+1119873))
1 + ℎ(120574 + 119889 + 120576 minus 120573(119878119894+1119873))
120582119899minus119894minus1
3=120582119899minus119894
3minus 1198603ℎ
1 + 119889ℎ
119879119894+1=(120582119899minus119894minus1+119898
1minus 120582119899minus119894minus1+119898
3)
1198604
120594[0119905119891minus120591]
(119905119894+1)119878119894+1
119906119894+1= min (119887max (0 119879119894+1))
Step 3 For 119894 = 0 n write119878lowast(119905119894) = 119878119894 119868lowast(119905119894) = 119868119894 119877lowast(119905
119894) = 119877119894 119906lowast(119905
119894) = 119906119894
End for
Algorithm 1
0
2
0 30 60 90 120 150 1800
5
Time (days)
times104 times10
6
With
out c
ontro
lI(t)
With
cont
rol
I(t)
Figure 2 The function 119868 with and without control
0 30 60 90 120 150 1800
05
1
15
2
Time (days)
times104
Infe
cted
I(t)
Figure 3 The function 119868 with control
Figures 2 3 and 4 represent the number of infectedindividuals (119868) with control (solid curve) and without control
0 30 60 90 120 150 1800
051
152
253
354
455
Time (days)
times106
Infe
cted
I(t)
Figure 4 The function 119868 without control
(dashed curve) It shows that in the presence of a controlthe number of infected individuals (119868) decreases greatly Themaximum number of infected individuals in the case withcontrol is 1787 times 104 and is 4404 times 106 in the case withoutcontrol then the efficiency of our strategy in reducing thespread of infection is nearly 9952
Figure 5 shows that the number of people removed beginsto grow notably from 24th day instead of 48th day inthe absence of control Moreover the number of recoveredindividuals at the end of the vaccination period is 2926 times107 instead of 2342 times 107 which represents an increase of
5840 times 106 cases
Figures 6 and 7 represent respectively the optimal con-trol and optimal value of the costThe curves start to increaseduring the first month because of the high infection level
6 ISRN Biomathematics
0 30 60 90 120 150 1800
05
1
15
2
25
3
Time (days)
times107
Rem
oved
R(t)
With controlWithout control
Figure 5 The function 119877 with and without control
0 30 60 90 120 150 1800
001
002
003
004
005
006
Time (days)
Con
trol
u(t)
Figure 6 The optimal control 119906lowast
0 30 60 90 120 150 1800
05
1
15
Time (days)
Cos
t
times10minus3
Figure 7 The optimal value of the cost
then they drop off steadily which is because of the constantand steady eradication of the infection
5 Conclusion
The work in this paper contributes to a growing literatureon applying optimal control techniques to epidemiology Weproposed amore realistic controlledmodel by including timedelay which represents the needed time for the migrationfrom the susceptible class to the recovered class after vac-cination The optimal control theory has been applied inthe context of an SIR model with time delay in state andcontrol variables and that includes a control 119906 that representsthe percentage of susceptible individuals being vaccinatedper time unit By using Pontryaginrsquos maximum principlethe explicit expression of the optimal controls was obtainedSimulation results indicate that the proposed control strategyis effective in reducing the number of susceptible and infectedindividuals and maximizing the recovered individuals
Acknowledgment
The authors would like to thank the anonymous referee forhisher valuable comments on the first version of the paperwhich have led to an improvement in this paper Researchreported in this paper was supported by the MoroccanSystemsTheory Network
References
[1] R M Anderson and R M May Infectious Diseases of HumansDynamics and Control Oxford University Press Oxford UK1992
[2] O Diekmann and J A P Heesterbeek Mathematical Epidemi-ology of Infectious Diseases John Wiley amp Sons Chisteter UK2000
[3] F Brauer and C C CastilloMathematical Models in PopulationBiology and Epidemiology Springer New York NY USA 2000
[4] P Van Den Driessche ldquoSome epidemiological models withdelaysrdquo inDifferential Equations andApplications to Biology andto Industry (Claremont CA 1994) pp 507ndash520 World Scien-tific River Edge NJ USA 1996
[5] Q J A Khan and E V Krishnan ldquoAn epidemic model with atime delay in transmissionrdquo Applications of Mathematics vol48 no 3 pp 193ndash203 2003
[6] N Yoshida and T Hara ldquoGlobal stability of a delayed SIRepidemic model with density dependent birth and death ratesrdquoJournal of Computational and AppliedMathematics vol 201 no2 pp 339ndash347 2007
[7] R Xu Z Ma and Z Wang ldquoGlobal stability of a delayed SIRSepidemic model with saturation incidence and temporaryimmunityrdquo Computers and Mathematics with Applications vol59 no 9 pp 3211ndash3221 2010
[8] XMeng L Chen and BWu ldquoA delay SIR epidemicmodel withpulse vaccination and incubation timesrdquo Nonlinear AnalysisReal World Applications vol 11 no 1 pp 88ndash98 2010
[9] G Zaman Y H Kang and I H Jung ldquoOptimal treatment of anSIR epidemic model with time delayrdquo BioSystems vol 98 no 1pp 43ndash50 2009
[10] K Hattaf and N Yousfi ldquoOptimal control of a delayed HIVinfection model with immune response using an efficientnumerical methodrdquo ISRN Biomathematics vol 2012 Article ID215124 7 pages 2012
ISRN Biomathematics 7
[11] G Birkhoff and G C Rota Ordinary Differential EquationsJohn Wiley amp Sons New YorkNY USA 4th edition 1989
[12] W H Fleming and R W Rishel Deterministic and StochasticOptimal Control Springer New York NY USA 1975
[13] D L Lukes Differential Equations Classical to Controlled Vol162 of Mathematics in Science and Engineering Academic PressNew York NY USA 1982
[14] L Gollmann D Kern and H Maurer ldquoOptimal control prob-lems with delays in state and control variables subject to mixedcontrol-state constraintsrdquo Optimal Control Applications andMethods vol 30 no 4 pp 341ndash365 2009
[15] K Hattaf and N Yousfi ldquoMathematical Model of the InfluenzaA(H1N1) Infectionrdquo Advanced Studies in Biology vol 1 no 8pp 383ndash390 2009
[16] M El hia O Balatif J Bouyaghroumni E Labriji and MRachik ldquoOptimal control applied to the spread of influenzaA(H1N1)rdquo Applied Mathematical Sciences vol 6 no 82 pp4057ndash4065 2012
[17] httpwwwpoumoncadiseases-maladiesa-zswineflu-grippe-porcinevaccines-vaccins fphp
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
2 ISRN Biomathematics
Section 3 In Section 4 we will give a numerical appropriatemethod and the corresponding simulation results Finally theconclusions are summarized in Section 5
2 Mathematical Model
We consider the SIR epidemic model with constant totalpopulation sizeThe population is divided into three disease-state compartments susceptible individuals (119878) people whocan catch the disease infectious (infective) individuals (119868)people who have the disease and can transmit the diseaserecovered individuals (119877) people who have recovered fromthe disease We assume that an individual can be infectedonly through contacts with infectious individuals and thatimmunity is permanent The transitions between differentstates are described by the following parameters
(i) Λ is the recruitment rate of susceptibles
(ii) 120573 is the effective contact rate
(iii) 119889 is the natural mortality rate
(iv) 120574 is the recovery rate
(v) 120598 is the disease induced death rate
The population dynamics is given by the following system ofordinary differential equations subject to nonnegative initialconditions
119889119878 (119905)
119889119905= Λ minus 120573119878 (119905)
119868 (119905)
119873 (119905)minus 119889119878 (119905)
119889119868 (119905)
119889119905= 120573119878 (119905)
119868 (119905)
119873 (119905)minus (120574 + 119889 + 120598) 119868 (119905)
119889119877 (119905)
119889119905= 120574119868 (119905) minus 119889119877 (119905)
(1)
where 119878(0) = 1198780 119868(0) = 119868
0119877(0) = 119877
0 and119873(119905) = 119878(119905)+119868(119905)+
119877(119905) is the total population number at time 119905The strategy of the control we adopt consists of a vaccina-
tion program our goal is to minimize the level of susceptibleand infected individuals and to maximize the recoveredindividuals Into the model (1) we include a control 119906 thatrepresents the percentage of susceptible individuals beingvaccinated per time unit In order to have a realisticmodel weneed to take into account that themovement of the vaccinatedsusceptible individuals from the class of susceptibles into therecovered class is subject to delay Thus the time delay isintroduced in the system as follows at time 119905 only a per-centage of susceptible individuals that have been vaccinated120591 time unit ago that is at time 119905 minus 120591 are removed from thesusceptible class and added to the recovered class So themathematical system with time delay in state and control
variables is given by the nonlinear retarded differentialequations
119889119878 (119905)
119889119905= Λ minus 120573119878 (119905)
119868 (119905)
119873 (119905)minus 119889119878 (119905) minus 119906 (119905 minus 120591) 119878 (119905 minus 120591)
119889119868 (119905)
119889119905= 120573119878 (119905)
119868 (119905)
119873 (119905)minus (120574 + 119889 + 120598) 119868 (119905)
119889119877 (119905)
119889119905= 120574119868 (119905) minus 119889119877 (119905) + 119906 (119905 minus 120591) 119878 (119905 minus 120591)
(2)
In addition for biological reasons we assume for 120579 isin [minus120591 0]that 119878(120579) 119868(120579) and 119877(120579) are nonnegative continuous func-tions and 119906(120579) = 0 Note that the control 119906 is assumed to beintegrable in the sense of Lebesgue bounded with 0 le 119906 le119887 lt 1 and 119887 is a given constant
To show the existence of solutions for the control system(2) we first prove that the system (2) is dissipative that is allsolutions are uniformly bounded in a proper subset Ω isin R3
+
Let (119878 119868 119877) isin R3+be any solution with nonnegative initial
conditions Adding equations of (2) we get
119889119873
119889119905= Λ minus 119889119873 minus 120598119868 lt Λ minus 119889119873 (3)
After integration using the constant variation formula wehave
119873(119905) leΛ
119889+ 119873 (0) 119890
minus119889119905 (4)
It then follows that
0 le 119873 (119905) leΛ
119889as 119905 997888rarr infin (5)
Therefore all feasible solutions of the system (2) enter into theregion
Ω = (119878 119868 119877) isin R3
+ 119873 le
Λ
119889 (6)
Then we can rewrite (2) in the following form
119889119883
119889119905= 119860119883 + 119865 (119883119883
120591) = 119866 (119883119883
120591) (7)
where
119883 (119905) = [
[
119878 (119905)
119868 (119905)
119877 (119905)
]
]
119860 = [
[
minus119889 0 0
0 minus (119889 + 120574 + 120576) 0
0 120574 minus119889
]
]
119865 (119883 (119905) 119883120591(119905)) =
[[[[[[
[
Λ minus120573119878 (119905) 119868 (119905)
119873 (119905)minus 119906120591(119905) 119878120591(119905)
120573119878 (119905) 119868 (119905)
119873 (119905)
119906120591 (119905) 119878120591 (119905)
]]]]]]
]
119906120591(119905) = 119906 (119905 minus 120591) 119883
120591(119905) = 119883 (119905 minus 120591)
(8)
ISRN Biomathematics 3
The second term on the right-hand side of (7) satisfies1003816100381610038161003816119865 (1198831 (119905) 1198831120591 (119905)) minus 119865 (1198832 (119905) 1198832120591 (119905))
1003816100381610038161003816
le 1198721
10038161003816100381610038161198831 (119905) minus 1198832 (119905)1003816100381610038161003816
+ 1198722
10038161003816100381610038161198831120591 (119905) minus 1198832120591 (119905)1003816100381610038161003816
(9)
where1198721and119872
2are some positive constants independent
of the state variables 119878(119905) 119868(119905) and 119877(119905) and10038161003816100381610038161198831 (119905) minus 1198832 (119905)
1003816100381610038161003816 =10038161003816100381610038161198781 (119905) minus 1198782 (119905)
1003816100381610038161003816
+10038161003816100381610038161198681 (119905) minus 1198682 (119905)
1003816100381610038161003816 +10038161003816100381610038161198771 (119905) minus 1198772 (119905)
1003816100381610038161003816
10038161003816100381610038161198831120591 (119905) minus 1198832120591 (119905)1003816100381610038161003816 =
10038161003816100381610038161198781120591 (119905) minus 1198782120591 (119905)1003816100381610038161003816
+10038161003816100381610038161198681120591 (119905) minus 1198682120591 (119905)
1003816100381610038161003816 +10038161003816100381610038161198771120591 (119905) minus 1198772120591 (119905)
1003816100381610038161003816
(10)
Here
119878119894120591 (119905) = 119878119894 (119905 minus 120591) 119868
119894120591 (119905) = 119868119894 (119905 minus 120591)
119877119894120591(119905) = 119877
119894(119905 minus 120591) for 119894 = 1 2
(11)
Moreover we get1003816100381610038161003816119866 (1198831 1198831120591) minus 119866 (1198832 1198832120591)
1003816100381610038161003816 le 119872 (10038161003816100381610038161198831 minus 1198832
1003816100381610038161003816 +10038161003816100381610038161198831120591 minus 1198832120591
1003816100381610038161003816)
(12)
where
119872 = max (1198721+ 119860 1198722) lt infin (13)
Thus it follows that the function119866 is uniformly Lipschitzcontinuous From the definition of the control 119906(119905) and therestriction on 119878(119905) 119868(119905) and 119877(119905) ge 0 we see that a solutionof the system (7) exists (see [11])
3 The Optimal Control Problem
Our goal is to minimize the number of susceptible andinfected individuals and to maximize the number of recov-ered individuals during the course of an epidemic Mathe-matically for a fixed terminal time 119905
119891 the problem is to min-
imize the objective functional
119869 (119906) = int
119905119891
0
1198601119878 (119905) + 119860
2119868 (119905) minus 119860
3119877 (119905) +
1198604
21199062(119905) 119889119905
(14)
where 119860119894ge 0 (for 119894 = 1 4) denote weights that balance
the size of the terms In other words we seek the optimalcontrol 119906lowast such that
119869 (119906lowast) = min 119869 (119906) 119906 isin U (15)
whereU is the set of admissible controls defined by
U = 119906 0 le 119906 le 119887 lt 1 119905 isin [0 119905119891]
119906 is Lebesgue measurable (16)
31 Existence of an Optimal Control The existence of theoptimal control can be obtained using a result by Fleming andRishel in [12]
Theorem 1 Consider the control problem with system (2)There exists an optimal control 119906lowast isin U such that
119869 (119906lowast) = min119906isinU
119869 (119906) (17)
Proof To use an existence result in [12] we must check thefollowing properties
(1) The set of controls and corresponding state variablesis nonempty
(2) The control setU is convex and closed
(3) The right-hand side of the state system is bounded bya linear function in the state and control variables
(4) The integrand of the objective functional is convex onU
(5) There exist constants 1198881 1198882gt 0 and 120588 gt 1 such that
the integrand 119871(119878 119868 119877 119906) of the objective functionalsatisfies
119871 (119878 119868 119877 119906) ge 1198882+ 1198881(|119906|2)1205882
(18)
An existence result by Lukes [13] was used to give the exis-tence of solution of system (2) with bounded coefficientswhich gives condition 1 The control set is convex and closedby definition Since the state system is linear in 119906 the rightside of (2) satisfies condition 3 using the boundedness ofthe solution The integrand in the objective functional (14) isconvex onU In addition we can easily see that there exist aconstant 120588 gt 1 and positive numbers 119888
1and 1198882satisfying
119871 (119878 119868 119877 119906) ge 1198882+ 1198881(|119906|2)1205882
(19)
32 Characterization of the Optimal Control In order toderive the necessary condition for the optimal controlPontryaginrsquos maximum principal with delay given in [14] wasusedThis principal converts (2) (14) and (15) into a problemof minimizing a Hamiltonian119867 defined by
119867 = 1198601119878 (119905) + 119860
2119868 (119905) minus 119860
3119877 (119905) +
1198604
21199062(119905) +
3
sum
119894=1
120582119894119891119894
(20)
where 119891119894is the right side of the differential equation of the 119894th
state variable
4 ISRN Biomathematics
Theorem 2 Given an optimal control 119906lowast isin U and solutions119878lowast 119868lowast and 119877lowast of the corresponding state system (2) there existadjoint functions 120582
1 1205822 and 120582
3satisfying
1= minus 119860
1+ 1205821119889 + (120582
1minus 1205822) 120573
119868lowast
119873lowast
+ 120594[0119905119891minus120591]
(119905) (120582+
1minus 120582+
3) 119906lowast
2= minus119860
2+ (1205821minus 1205822) 120573
119878lowast
119873lowast+ (119889 + 120574 + 120598) 120582
2minus 1205741205823
3= 1198603+ 1205823119889
(21)
with the transversality conditions
1205821(119905119891) = 1205822(119905119891) = 1205823(119905119891) = 0 (22)
Furthermore the optimal control 119906lowast is given by
119906lowast(119905) = min(119887max(0
(120582+
1minus 120582+
3)
1198604
120594[0119905119891minus120591]
(119905) 119878lowast)) (23)
where 120582+119894(119905) = 120582
119894(119905 + 120591) for 119894 = 1 3
Proof The adjoint equations and transversality conditionscan be obtained by using Pontryaginrsquos maximum principlewith delay in the state and control variables [14] such that
1= minus
120597119867 (119905)
120597119878 (119905)minus 120594[0119905119891minus120591]
(119905)120597119867 (119905 + 120591)
120597119878 (119905 minus 120591) 120582
1(119905119891) = 0
2= minus
120597119867
120597119868 120582
2(119905119891) = 0
3= minus
120597119867
120597119877 120582
3(119905119891) = 0
(24)
The optimal control 119906lowast can be solved from the optimalitycondition
120597119867
120597119906+ 120594[0119905119891minus120591]
(119905)120597119867 (119905 + 120591)
120597119906120591
= 0 (25)
That is
1198604119906 + 120594[0119905119891minus120591]
(119905) (120582+
3minus 120582+
1) 119878 = 0 (26)
By the bounds in U of the control it is easy to rewrite 119906lowast inthe form (23)
4 Numerical Simulation
In this section we give a numerical method to solve the opti-mality system which is a two-point boundary value problemwith separated boundary conditions at times 119905
0= 0 and 119905
119891
Let there exist a step size ℎ gt 0 and (119899119898) isin N2 with120591 = 119898ℎ and 119905
119891minus 1199050= 119899ℎ For reasons of programming we
0 30 60 90 120 150 1800
05
1
15
2
25
3
Time (days)
With controlWithout control
Susc
eptib
les
S(t)
times107
Figure 1 The function 119878 with and without control
consider 119898 knots to left of 1199050and right of 119905
119891 and we obtain
the following partition
Δ = (119905minus119898= minus120591 lt sdot sdot sdot lt 119905
minus1lt 1199050= 0 lt 119905
1lt sdot sdot sdot lt 119905
119899
= 119905119891lt 119905119899+1
lt sdot sdot sdot lt 119905119899+119898
)
(27)
Then we have 119905119894= 1199050+ 119894ℎ (minus119898 le 119894 le 119899 + 119898) Next we
define the state and adjoint variables 119878(119905) 119868(119905) 119877(119905) 1205821 1205822
1205823 and the control 119906 in terms of nodal points 119878
119894 119868119894 119877119894 1205821198941 1205821198942
120582119894
3 and 119906
119894 Now using combination of forward and backward
difference approximations we obtain the Algorithm 1The numerical simulations were carried out using data
regarding the course of the influenza A (H1N1) in MoroccoThe initial conditions and parameters of the system (2) aretaken from [15 16] while the time delay value is taken from[17]
120573 = 03095 Λ = 117417
119889 = 39139 times 10minus5 120574 = 02
120576 = 00063 120591 = 10
(28)
the initial conditions for the ordinary differential systemwere
119878 (0) = 30 times 106 119868 (0) = 30 119877 (0) = 28 (29)
and the transversality conditions for the ordinary differentialsystem were
120582119894(119905119891) = 0 (119894 = 1 3) (30)
Figure 1 indicates that the number of susceptible indi-viduals (119878) decreases more rapidly in the case with controlIt reaches 7399 times 105 at the end of the vaccination periodagainst 6732 times 106 in case without control that is a reduc-tion of 5992 times 106
ISRN Biomathematics 5
Step 1 For 119894 = minus119898 0 do 119878119894= 1198780 119868119894= 1198680 119877119894= 1198770 119906119894= 0
End forFor 119894 = 119899 119899 + 119898 do 120582119894
1= 0 120582119894
2= 0 120582119894
3= 0
End forStep 2 For 119894 = 0 119899 minus 1 do
119878119894+1=119878119894+ ℎ(Λ minus 119906
119894minus119898119878119894minus119898)
1 + ℎ(120573(119868119894119873) + 119889)
119868119894+1=
119868119894
1 + ℎ(120574 + 119889 + 120576 minus 120573(119878119894+1119873))
119877119894+1=119877119894+ ℎ(120574119868
119894+1+ 119906119894minus119898119878119894minus119898)
1 + ℎ119889
120582119899minus119894minus1
1=
120582119899minus119894
1+ ℎ(119860
1+ 120573(119868
119894+1119873)120582
119899minus119894
2+ (120582119899minus119894+119898
3minus 120582119899minus119894minus1+119898
1)120594[0119905119891minus120591]
(119905119899minus119894)119906119894)
1 + ℎ(119889 + 120573(119868119894+1119873))
120582119899minus119894minus1
2=120582119899minus119894
2+ ℎ(119860
2+ 120582119899minus119894
3120574 minus 120582119899minus119894minus1
1120573(119878119894+1119873))
1 + ℎ(120574 + 119889 + 120576 minus 120573(119878119894+1119873))
120582119899minus119894minus1
3=120582119899minus119894
3minus 1198603ℎ
1 + 119889ℎ
119879119894+1=(120582119899minus119894minus1+119898
1minus 120582119899minus119894minus1+119898
3)
1198604
120594[0119905119891minus120591]
(119905119894+1)119878119894+1
119906119894+1= min (119887max (0 119879119894+1))
Step 3 For 119894 = 0 n write119878lowast(119905119894) = 119878119894 119868lowast(119905119894) = 119868119894 119877lowast(119905
119894) = 119877119894 119906lowast(119905
119894) = 119906119894
End for
Algorithm 1
0
2
0 30 60 90 120 150 1800
5
Time (days)
times104 times10
6
With
out c
ontro
lI(t)
With
cont
rol
I(t)
Figure 2 The function 119868 with and without control
0 30 60 90 120 150 1800
05
1
15
2
Time (days)
times104
Infe
cted
I(t)
Figure 3 The function 119868 with control
Figures 2 3 and 4 represent the number of infectedindividuals (119868) with control (solid curve) and without control
0 30 60 90 120 150 1800
051
152
253
354
455
Time (days)
times106
Infe
cted
I(t)
Figure 4 The function 119868 without control
(dashed curve) It shows that in the presence of a controlthe number of infected individuals (119868) decreases greatly Themaximum number of infected individuals in the case withcontrol is 1787 times 104 and is 4404 times 106 in the case withoutcontrol then the efficiency of our strategy in reducing thespread of infection is nearly 9952
Figure 5 shows that the number of people removed beginsto grow notably from 24th day instead of 48th day inthe absence of control Moreover the number of recoveredindividuals at the end of the vaccination period is 2926 times107 instead of 2342 times 107 which represents an increase of
5840 times 106 cases
Figures 6 and 7 represent respectively the optimal con-trol and optimal value of the costThe curves start to increaseduring the first month because of the high infection level
6 ISRN Biomathematics
0 30 60 90 120 150 1800
05
1
15
2
25
3
Time (days)
times107
Rem
oved
R(t)
With controlWithout control
Figure 5 The function 119877 with and without control
0 30 60 90 120 150 1800
001
002
003
004
005
006
Time (days)
Con
trol
u(t)
Figure 6 The optimal control 119906lowast
0 30 60 90 120 150 1800
05
1
15
Time (days)
Cos
t
times10minus3
Figure 7 The optimal value of the cost
then they drop off steadily which is because of the constantand steady eradication of the infection
5 Conclusion
The work in this paper contributes to a growing literatureon applying optimal control techniques to epidemiology Weproposed amore realistic controlledmodel by including timedelay which represents the needed time for the migrationfrom the susceptible class to the recovered class after vac-cination The optimal control theory has been applied inthe context of an SIR model with time delay in state andcontrol variables and that includes a control 119906 that representsthe percentage of susceptible individuals being vaccinatedper time unit By using Pontryaginrsquos maximum principlethe explicit expression of the optimal controls was obtainedSimulation results indicate that the proposed control strategyis effective in reducing the number of susceptible and infectedindividuals and maximizing the recovered individuals
Acknowledgment
The authors would like to thank the anonymous referee forhisher valuable comments on the first version of the paperwhich have led to an improvement in this paper Researchreported in this paper was supported by the MoroccanSystemsTheory Network
References
[1] R M Anderson and R M May Infectious Diseases of HumansDynamics and Control Oxford University Press Oxford UK1992
[2] O Diekmann and J A P Heesterbeek Mathematical Epidemi-ology of Infectious Diseases John Wiley amp Sons Chisteter UK2000
[3] F Brauer and C C CastilloMathematical Models in PopulationBiology and Epidemiology Springer New York NY USA 2000
[4] P Van Den Driessche ldquoSome epidemiological models withdelaysrdquo inDifferential Equations andApplications to Biology andto Industry (Claremont CA 1994) pp 507ndash520 World Scien-tific River Edge NJ USA 1996
[5] Q J A Khan and E V Krishnan ldquoAn epidemic model with atime delay in transmissionrdquo Applications of Mathematics vol48 no 3 pp 193ndash203 2003
[6] N Yoshida and T Hara ldquoGlobal stability of a delayed SIRepidemic model with density dependent birth and death ratesrdquoJournal of Computational and AppliedMathematics vol 201 no2 pp 339ndash347 2007
[7] R Xu Z Ma and Z Wang ldquoGlobal stability of a delayed SIRSepidemic model with saturation incidence and temporaryimmunityrdquo Computers and Mathematics with Applications vol59 no 9 pp 3211ndash3221 2010
[8] XMeng L Chen and BWu ldquoA delay SIR epidemicmodel withpulse vaccination and incubation timesrdquo Nonlinear AnalysisReal World Applications vol 11 no 1 pp 88ndash98 2010
[9] G Zaman Y H Kang and I H Jung ldquoOptimal treatment of anSIR epidemic model with time delayrdquo BioSystems vol 98 no 1pp 43ndash50 2009
[10] K Hattaf and N Yousfi ldquoOptimal control of a delayed HIVinfection model with immune response using an efficientnumerical methodrdquo ISRN Biomathematics vol 2012 Article ID215124 7 pages 2012
ISRN Biomathematics 7
[11] G Birkhoff and G C Rota Ordinary Differential EquationsJohn Wiley amp Sons New YorkNY USA 4th edition 1989
[12] W H Fleming and R W Rishel Deterministic and StochasticOptimal Control Springer New York NY USA 1975
[13] D L Lukes Differential Equations Classical to Controlled Vol162 of Mathematics in Science and Engineering Academic PressNew York NY USA 1982
[14] L Gollmann D Kern and H Maurer ldquoOptimal control prob-lems with delays in state and control variables subject to mixedcontrol-state constraintsrdquo Optimal Control Applications andMethods vol 30 no 4 pp 341ndash365 2009
[15] K Hattaf and N Yousfi ldquoMathematical Model of the InfluenzaA(H1N1) Infectionrdquo Advanced Studies in Biology vol 1 no 8pp 383ndash390 2009
[16] M El hia O Balatif J Bouyaghroumni E Labriji and MRachik ldquoOptimal control applied to the spread of influenzaA(H1N1)rdquo Applied Mathematical Sciences vol 6 no 82 pp4057ndash4065 2012
[17] httpwwwpoumoncadiseases-maladiesa-zswineflu-grippe-porcinevaccines-vaccins fphp
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
ISRN Biomathematics 3
The second term on the right-hand side of (7) satisfies1003816100381610038161003816119865 (1198831 (119905) 1198831120591 (119905)) minus 119865 (1198832 (119905) 1198832120591 (119905))
1003816100381610038161003816
le 1198721
10038161003816100381610038161198831 (119905) minus 1198832 (119905)1003816100381610038161003816
+ 1198722
10038161003816100381610038161198831120591 (119905) minus 1198832120591 (119905)1003816100381610038161003816
(9)
where1198721and119872
2are some positive constants independent
of the state variables 119878(119905) 119868(119905) and 119877(119905) and10038161003816100381610038161198831 (119905) minus 1198832 (119905)
1003816100381610038161003816 =10038161003816100381610038161198781 (119905) minus 1198782 (119905)
1003816100381610038161003816
+10038161003816100381610038161198681 (119905) minus 1198682 (119905)
1003816100381610038161003816 +10038161003816100381610038161198771 (119905) minus 1198772 (119905)
1003816100381610038161003816
10038161003816100381610038161198831120591 (119905) minus 1198832120591 (119905)1003816100381610038161003816 =
10038161003816100381610038161198781120591 (119905) minus 1198782120591 (119905)1003816100381610038161003816
+10038161003816100381610038161198681120591 (119905) minus 1198682120591 (119905)
1003816100381610038161003816 +10038161003816100381610038161198771120591 (119905) minus 1198772120591 (119905)
1003816100381610038161003816
(10)
Here
119878119894120591 (119905) = 119878119894 (119905 minus 120591) 119868
119894120591 (119905) = 119868119894 (119905 minus 120591)
119877119894120591(119905) = 119877
119894(119905 minus 120591) for 119894 = 1 2
(11)
Moreover we get1003816100381610038161003816119866 (1198831 1198831120591) minus 119866 (1198832 1198832120591)
1003816100381610038161003816 le 119872 (10038161003816100381610038161198831 minus 1198832
1003816100381610038161003816 +10038161003816100381610038161198831120591 minus 1198832120591
1003816100381610038161003816)
(12)
where
119872 = max (1198721+ 119860 1198722) lt infin (13)
Thus it follows that the function119866 is uniformly Lipschitzcontinuous From the definition of the control 119906(119905) and therestriction on 119878(119905) 119868(119905) and 119877(119905) ge 0 we see that a solutionof the system (7) exists (see [11])
3 The Optimal Control Problem
Our goal is to minimize the number of susceptible andinfected individuals and to maximize the number of recov-ered individuals during the course of an epidemic Mathe-matically for a fixed terminal time 119905
119891 the problem is to min-
imize the objective functional
119869 (119906) = int
119905119891
0
1198601119878 (119905) + 119860
2119868 (119905) minus 119860
3119877 (119905) +
1198604
21199062(119905) 119889119905
(14)
where 119860119894ge 0 (for 119894 = 1 4) denote weights that balance
the size of the terms In other words we seek the optimalcontrol 119906lowast such that
119869 (119906lowast) = min 119869 (119906) 119906 isin U (15)
whereU is the set of admissible controls defined by
U = 119906 0 le 119906 le 119887 lt 1 119905 isin [0 119905119891]
119906 is Lebesgue measurable (16)
31 Existence of an Optimal Control The existence of theoptimal control can be obtained using a result by Fleming andRishel in [12]
Theorem 1 Consider the control problem with system (2)There exists an optimal control 119906lowast isin U such that
119869 (119906lowast) = min119906isinU
119869 (119906) (17)
Proof To use an existence result in [12] we must check thefollowing properties
(1) The set of controls and corresponding state variablesis nonempty
(2) The control setU is convex and closed
(3) The right-hand side of the state system is bounded bya linear function in the state and control variables
(4) The integrand of the objective functional is convex onU
(5) There exist constants 1198881 1198882gt 0 and 120588 gt 1 such that
the integrand 119871(119878 119868 119877 119906) of the objective functionalsatisfies
119871 (119878 119868 119877 119906) ge 1198882+ 1198881(|119906|2)1205882
(18)
An existence result by Lukes [13] was used to give the exis-tence of solution of system (2) with bounded coefficientswhich gives condition 1 The control set is convex and closedby definition Since the state system is linear in 119906 the rightside of (2) satisfies condition 3 using the boundedness ofthe solution The integrand in the objective functional (14) isconvex onU In addition we can easily see that there exist aconstant 120588 gt 1 and positive numbers 119888
1and 1198882satisfying
119871 (119878 119868 119877 119906) ge 1198882+ 1198881(|119906|2)1205882
(19)
32 Characterization of the Optimal Control In order toderive the necessary condition for the optimal controlPontryaginrsquos maximum principal with delay given in [14] wasusedThis principal converts (2) (14) and (15) into a problemof minimizing a Hamiltonian119867 defined by
119867 = 1198601119878 (119905) + 119860
2119868 (119905) minus 119860
3119877 (119905) +
1198604
21199062(119905) +
3
sum
119894=1
120582119894119891119894
(20)
where 119891119894is the right side of the differential equation of the 119894th
state variable
4 ISRN Biomathematics
Theorem 2 Given an optimal control 119906lowast isin U and solutions119878lowast 119868lowast and 119877lowast of the corresponding state system (2) there existadjoint functions 120582
1 1205822 and 120582
3satisfying
1= minus 119860
1+ 1205821119889 + (120582
1minus 1205822) 120573
119868lowast
119873lowast
+ 120594[0119905119891minus120591]
(119905) (120582+
1minus 120582+
3) 119906lowast
2= minus119860
2+ (1205821minus 1205822) 120573
119878lowast
119873lowast+ (119889 + 120574 + 120598) 120582
2minus 1205741205823
3= 1198603+ 1205823119889
(21)
with the transversality conditions
1205821(119905119891) = 1205822(119905119891) = 1205823(119905119891) = 0 (22)
Furthermore the optimal control 119906lowast is given by
119906lowast(119905) = min(119887max(0
(120582+
1minus 120582+
3)
1198604
120594[0119905119891minus120591]
(119905) 119878lowast)) (23)
where 120582+119894(119905) = 120582
119894(119905 + 120591) for 119894 = 1 3
Proof The adjoint equations and transversality conditionscan be obtained by using Pontryaginrsquos maximum principlewith delay in the state and control variables [14] such that
1= minus
120597119867 (119905)
120597119878 (119905)minus 120594[0119905119891minus120591]
(119905)120597119867 (119905 + 120591)
120597119878 (119905 minus 120591) 120582
1(119905119891) = 0
2= minus
120597119867
120597119868 120582
2(119905119891) = 0
3= minus
120597119867
120597119877 120582
3(119905119891) = 0
(24)
The optimal control 119906lowast can be solved from the optimalitycondition
120597119867
120597119906+ 120594[0119905119891minus120591]
(119905)120597119867 (119905 + 120591)
120597119906120591
= 0 (25)
That is
1198604119906 + 120594[0119905119891minus120591]
(119905) (120582+
3minus 120582+
1) 119878 = 0 (26)
By the bounds in U of the control it is easy to rewrite 119906lowast inthe form (23)
4 Numerical Simulation
In this section we give a numerical method to solve the opti-mality system which is a two-point boundary value problemwith separated boundary conditions at times 119905
0= 0 and 119905
119891
Let there exist a step size ℎ gt 0 and (119899119898) isin N2 with120591 = 119898ℎ and 119905
119891minus 1199050= 119899ℎ For reasons of programming we
0 30 60 90 120 150 1800
05
1
15
2
25
3
Time (days)
With controlWithout control
Susc
eptib
les
S(t)
times107
Figure 1 The function 119878 with and without control
consider 119898 knots to left of 1199050and right of 119905
119891 and we obtain
the following partition
Δ = (119905minus119898= minus120591 lt sdot sdot sdot lt 119905
minus1lt 1199050= 0 lt 119905
1lt sdot sdot sdot lt 119905
119899
= 119905119891lt 119905119899+1
lt sdot sdot sdot lt 119905119899+119898
)
(27)
Then we have 119905119894= 1199050+ 119894ℎ (minus119898 le 119894 le 119899 + 119898) Next we
define the state and adjoint variables 119878(119905) 119868(119905) 119877(119905) 1205821 1205822
1205823 and the control 119906 in terms of nodal points 119878
119894 119868119894 119877119894 1205821198941 1205821198942
120582119894
3 and 119906
119894 Now using combination of forward and backward
difference approximations we obtain the Algorithm 1The numerical simulations were carried out using data
regarding the course of the influenza A (H1N1) in MoroccoThe initial conditions and parameters of the system (2) aretaken from [15 16] while the time delay value is taken from[17]
120573 = 03095 Λ = 117417
119889 = 39139 times 10minus5 120574 = 02
120576 = 00063 120591 = 10
(28)
the initial conditions for the ordinary differential systemwere
119878 (0) = 30 times 106 119868 (0) = 30 119877 (0) = 28 (29)
and the transversality conditions for the ordinary differentialsystem were
120582119894(119905119891) = 0 (119894 = 1 3) (30)
Figure 1 indicates that the number of susceptible indi-viduals (119878) decreases more rapidly in the case with controlIt reaches 7399 times 105 at the end of the vaccination periodagainst 6732 times 106 in case without control that is a reduc-tion of 5992 times 106
ISRN Biomathematics 5
Step 1 For 119894 = minus119898 0 do 119878119894= 1198780 119868119894= 1198680 119877119894= 1198770 119906119894= 0
End forFor 119894 = 119899 119899 + 119898 do 120582119894
1= 0 120582119894
2= 0 120582119894
3= 0
End forStep 2 For 119894 = 0 119899 minus 1 do
119878119894+1=119878119894+ ℎ(Λ minus 119906
119894minus119898119878119894minus119898)
1 + ℎ(120573(119868119894119873) + 119889)
119868119894+1=
119868119894
1 + ℎ(120574 + 119889 + 120576 minus 120573(119878119894+1119873))
119877119894+1=119877119894+ ℎ(120574119868
119894+1+ 119906119894minus119898119878119894minus119898)
1 + ℎ119889
120582119899minus119894minus1
1=
120582119899minus119894
1+ ℎ(119860
1+ 120573(119868
119894+1119873)120582
119899minus119894
2+ (120582119899minus119894+119898
3minus 120582119899minus119894minus1+119898
1)120594[0119905119891minus120591]
(119905119899minus119894)119906119894)
1 + ℎ(119889 + 120573(119868119894+1119873))
120582119899minus119894minus1
2=120582119899minus119894
2+ ℎ(119860
2+ 120582119899minus119894
3120574 minus 120582119899minus119894minus1
1120573(119878119894+1119873))
1 + ℎ(120574 + 119889 + 120576 minus 120573(119878119894+1119873))
120582119899minus119894minus1
3=120582119899minus119894
3minus 1198603ℎ
1 + 119889ℎ
119879119894+1=(120582119899minus119894minus1+119898
1minus 120582119899minus119894minus1+119898
3)
1198604
120594[0119905119891minus120591]
(119905119894+1)119878119894+1
119906119894+1= min (119887max (0 119879119894+1))
Step 3 For 119894 = 0 n write119878lowast(119905119894) = 119878119894 119868lowast(119905119894) = 119868119894 119877lowast(119905
119894) = 119877119894 119906lowast(119905
119894) = 119906119894
End for
Algorithm 1
0
2
0 30 60 90 120 150 1800
5
Time (days)
times104 times10
6
With
out c
ontro
lI(t)
With
cont
rol
I(t)
Figure 2 The function 119868 with and without control
0 30 60 90 120 150 1800
05
1
15
2
Time (days)
times104
Infe
cted
I(t)
Figure 3 The function 119868 with control
Figures 2 3 and 4 represent the number of infectedindividuals (119868) with control (solid curve) and without control
0 30 60 90 120 150 1800
051
152
253
354
455
Time (days)
times106
Infe
cted
I(t)
Figure 4 The function 119868 without control
(dashed curve) It shows that in the presence of a controlthe number of infected individuals (119868) decreases greatly Themaximum number of infected individuals in the case withcontrol is 1787 times 104 and is 4404 times 106 in the case withoutcontrol then the efficiency of our strategy in reducing thespread of infection is nearly 9952
Figure 5 shows that the number of people removed beginsto grow notably from 24th day instead of 48th day inthe absence of control Moreover the number of recoveredindividuals at the end of the vaccination period is 2926 times107 instead of 2342 times 107 which represents an increase of
5840 times 106 cases
Figures 6 and 7 represent respectively the optimal con-trol and optimal value of the costThe curves start to increaseduring the first month because of the high infection level
6 ISRN Biomathematics
0 30 60 90 120 150 1800
05
1
15
2
25
3
Time (days)
times107
Rem
oved
R(t)
With controlWithout control
Figure 5 The function 119877 with and without control
0 30 60 90 120 150 1800
001
002
003
004
005
006
Time (days)
Con
trol
u(t)
Figure 6 The optimal control 119906lowast
0 30 60 90 120 150 1800
05
1
15
Time (days)
Cos
t
times10minus3
Figure 7 The optimal value of the cost
then they drop off steadily which is because of the constantand steady eradication of the infection
5 Conclusion
The work in this paper contributes to a growing literatureon applying optimal control techniques to epidemiology Weproposed amore realistic controlledmodel by including timedelay which represents the needed time for the migrationfrom the susceptible class to the recovered class after vac-cination The optimal control theory has been applied inthe context of an SIR model with time delay in state andcontrol variables and that includes a control 119906 that representsthe percentage of susceptible individuals being vaccinatedper time unit By using Pontryaginrsquos maximum principlethe explicit expression of the optimal controls was obtainedSimulation results indicate that the proposed control strategyis effective in reducing the number of susceptible and infectedindividuals and maximizing the recovered individuals
Acknowledgment
The authors would like to thank the anonymous referee forhisher valuable comments on the first version of the paperwhich have led to an improvement in this paper Researchreported in this paper was supported by the MoroccanSystemsTheory Network
References
[1] R M Anderson and R M May Infectious Diseases of HumansDynamics and Control Oxford University Press Oxford UK1992
[2] O Diekmann and J A P Heesterbeek Mathematical Epidemi-ology of Infectious Diseases John Wiley amp Sons Chisteter UK2000
[3] F Brauer and C C CastilloMathematical Models in PopulationBiology and Epidemiology Springer New York NY USA 2000
[4] P Van Den Driessche ldquoSome epidemiological models withdelaysrdquo inDifferential Equations andApplications to Biology andto Industry (Claremont CA 1994) pp 507ndash520 World Scien-tific River Edge NJ USA 1996
[5] Q J A Khan and E V Krishnan ldquoAn epidemic model with atime delay in transmissionrdquo Applications of Mathematics vol48 no 3 pp 193ndash203 2003
[6] N Yoshida and T Hara ldquoGlobal stability of a delayed SIRepidemic model with density dependent birth and death ratesrdquoJournal of Computational and AppliedMathematics vol 201 no2 pp 339ndash347 2007
[7] R Xu Z Ma and Z Wang ldquoGlobal stability of a delayed SIRSepidemic model with saturation incidence and temporaryimmunityrdquo Computers and Mathematics with Applications vol59 no 9 pp 3211ndash3221 2010
[8] XMeng L Chen and BWu ldquoA delay SIR epidemicmodel withpulse vaccination and incubation timesrdquo Nonlinear AnalysisReal World Applications vol 11 no 1 pp 88ndash98 2010
[9] G Zaman Y H Kang and I H Jung ldquoOptimal treatment of anSIR epidemic model with time delayrdquo BioSystems vol 98 no 1pp 43ndash50 2009
[10] K Hattaf and N Yousfi ldquoOptimal control of a delayed HIVinfection model with immune response using an efficientnumerical methodrdquo ISRN Biomathematics vol 2012 Article ID215124 7 pages 2012
ISRN Biomathematics 7
[11] G Birkhoff and G C Rota Ordinary Differential EquationsJohn Wiley amp Sons New YorkNY USA 4th edition 1989
[12] W H Fleming and R W Rishel Deterministic and StochasticOptimal Control Springer New York NY USA 1975
[13] D L Lukes Differential Equations Classical to Controlled Vol162 of Mathematics in Science and Engineering Academic PressNew York NY USA 1982
[14] L Gollmann D Kern and H Maurer ldquoOptimal control prob-lems with delays in state and control variables subject to mixedcontrol-state constraintsrdquo Optimal Control Applications andMethods vol 30 no 4 pp 341ndash365 2009
[15] K Hattaf and N Yousfi ldquoMathematical Model of the InfluenzaA(H1N1) Infectionrdquo Advanced Studies in Biology vol 1 no 8pp 383ndash390 2009
[16] M El hia O Balatif J Bouyaghroumni E Labriji and MRachik ldquoOptimal control applied to the spread of influenzaA(H1N1)rdquo Applied Mathematical Sciences vol 6 no 82 pp4057ndash4065 2012
[17] httpwwwpoumoncadiseases-maladiesa-zswineflu-grippe-porcinevaccines-vaccins fphp
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
4 ISRN Biomathematics
Theorem 2 Given an optimal control 119906lowast isin U and solutions119878lowast 119868lowast and 119877lowast of the corresponding state system (2) there existadjoint functions 120582
1 1205822 and 120582
3satisfying
1= minus 119860
1+ 1205821119889 + (120582
1minus 1205822) 120573
119868lowast
119873lowast
+ 120594[0119905119891minus120591]
(119905) (120582+
1minus 120582+
3) 119906lowast
2= minus119860
2+ (1205821minus 1205822) 120573
119878lowast
119873lowast+ (119889 + 120574 + 120598) 120582
2minus 1205741205823
3= 1198603+ 1205823119889
(21)
with the transversality conditions
1205821(119905119891) = 1205822(119905119891) = 1205823(119905119891) = 0 (22)
Furthermore the optimal control 119906lowast is given by
119906lowast(119905) = min(119887max(0
(120582+
1minus 120582+
3)
1198604
120594[0119905119891minus120591]
(119905) 119878lowast)) (23)
where 120582+119894(119905) = 120582
119894(119905 + 120591) for 119894 = 1 3
Proof The adjoint equations and transversality conditionscan be obtained by using Pontryaginrsquos maximum principlewith delay in the state and control variables [14] such that
1= minus
120597119867 (119905)
120597119878 (119905)minus 120594[0119905119891minus120591]
(119905)120597119867 (119905 + 120591)
120597119878 (119905 minus 120591) 120582
1(119905119891) = 0
2= minus
120597119867
120597119868 120582
2(119905119891) = 0
3= minus
120597119867
120597119877 120582
3(119905119891) = 0
(24)
The optimal control 119906lowast can be solved from the optimalitycondition
120597119867
120597119906+ 120594[0119905119891minus120591]
(119905)120597119867 (119905 + 120591)
120597119906120591
= 0 (25)
That is
1198604119906 + 120594[0119905119891minus120591]
(119905) (120582+
3minus 120582+
1) 119878 = 0 (26)
By the bounds in U of the control it is easy to rewrite 119906lowast inthe form (23)
4 Numerical Simulation
In this section we give a numerical method to solve the opti-mality system which is a two-point boundary value problemwith separated boundary conditions at times 119905
0= 0 and 119905
119891
Let there exist a step size ℎ gt 0 and (119899119898) isin N2 with120591 = 119898ℎ and 119905
119891minus 1199050= 119899ℎ For reasons of programming we
0 30 60 90 120 150 1800
05
1
15
2
25
3
Time (days)
With controlWithout control
Susc
eptib
les
S(t)
times107
Figure 1 The function 119878 with and without control
consider 119898 knots to left of 1199050and right of 119905
119891 and we obtain
the following partition
Δ = (119905minus119898= minus120591 lt sdot sdot sdot lt 119905
minus1lt 1199050= 0 lt 119905
1lt sdot sdot sdot lt 119905
119899
= 119905119891lt 119905119899+1
lt sdot sdot sdot lt 119905119899+119898
)
(27)
Then we have 119905119894= 1199050+ 119894ℎ (minus119898 le 119894 le 119899 + 119898) Next we
define the state and adjoint variables 119878(119905) 119868(119905) 119877(119905) 1205821 1205822
1205823 and the control 119906 in terms of nodal points 119878
119894 119868119894 119877119894 1205821198941 1205821198942
120582119894
3 and 119906
119894 Now using combination of forward and backward
difference approximations we obtain the Algorithm 1The numerical simulations were carried out using data
regarding the course of the influenza A (H1N1) in MoroccoThe initial conditions and parameters of the system (2) aretaken from [15 16] while the time delay value is taken from[17]
120573 = 03095 Λ = 117417
119889 = 39139 times 10minus5 120574 = 02
120576 = 00063 120591 = 10
(28)
the initial conditions for the ordinary differential systemwere
119878 (0) = 30 times 106 119868 (0) = 30 119877 (0) = 28 (29)
and the transversality conditions for the ordinary differentialsystem were
120582119894(119905119891) = 0 (119894 = 1 3) (30)
Figure 1 indicates that the number of susceptible indi-viduals (119878) decreases more rapidly in the case with controlIt reaches 7399 times 105 at the end of the vaccination periodagainst 6732 times 106 in case without control that is a reduc-tion of 5992 times 106
ISRN Biomathematics 5
Step 1 For 119894 = minus119898 0 do 119878119894= 1198780 119868119894= 1198680 119877119894= 1198770 119906119894= 0
End forFor 119894 = 119899 119899 + 119898 do 120582119894
1= 0 120582119894
2= 0 120582119894
3= 0
End forStep 2 For 119894 = 0 119899 minus 1 do
119878119894+1=119878119894+ ℎ(Λ minus 119906
119894minus119898119878119894minus119898)
1 + ℎ(120573(119868119894119873) + 119889)
119868119894+1=
119868119894
1 + ℎ(120574 + 119889 + 120576 minus 120573(119878119894+1119873))
119877119894+1=119877119894+ ℎ(120574119868
119894+1+ 119906119894minus119898119878119894minus119898)
1 + ℎ119889
120582119899minus119894minus1
1=
120582119899minus119894
1+ ℎ(119860
1+ 120573(119868
119894+1119873)120582
119899minus119894
2+ (120582119899minus119894+119898
3minus 120582119899minus119894minus1+119898
1)120594[0119905119891minus120591]
(119905119899minus119894)119906119894)
1 + ℎ(119889 + 120573(119868119894+1119873))
120582119899minus119894minus1
2=120582119899minus119894
2+ ℎ(119860
2+ 120582119899minus119894
3120574 minus 120582119899minus119894minus1
1120573(119878119894+1119873))
1 + ℎ(120574 + 119889 + 120576 minus 120573(119878119894+1119873))
120582119899minus119894minus1
3=120582119899minus119894
3minus 1198603ℎ
1 + 119889ℎ
119879119894+1=(120582119899minus119894minus1+119898
1minus 120582119899minus119894minus1+119898
3)
1198604
120594[0119905119891minus120591]
(119905119894+1)119878119894+1
119906119894+1= min (119887max (0 119879119894+1))
Step 3 For 119894 = 0 n write119878lowast(119905119894) = 119878119894 119868lowast(119905119894) = 119868119894 119877lowast(119905
119894) = 119877119894 119906lowast(119905
119894) = 119906119894
End for
Algorithm 1
0
2
0 30 60 90 120 150 1800
5
Time (days)
times104 times10
6
With
out c
ontro
lI(t)
With
cont
rol
I(t)
Figure 2 The function 119868 with and without control
0 30 60 90 120 150 1800
05
1
15
2
Time (days)
times104
Infe
cted
I(t)
Figure 3 The function 119868 with control
Figures 2 3 and 4 represent the number of infectedindividuals (119868) with control (solid curve) and without control
0 30 60 90 120 150 1800
051
152
253
354
455
Time (days)
times106
Infe
cted
I(t)
Figure 4 The function 119868 without control
(dashed curve) It shows that in the presence of a controlthe number of infected individuals (119868) decreases greatly Themaximum number of infected individuals in the case withcontrol is 1787 times 104 and is 4404 times 106 in the case withoutcontrol then the efficiency of our strategy in reducing thespread of infection is nearly 9952
Figure 5 shows that the number of people removed beginsto grow notably from 24th day instead of 48th day inthe absence of control Moreover the number of recoveredindividuals at the end of the vaccination period is 2926 times107 instead of 2342 times 107 which represents an increase of
5840 times 106 cases
Figures 6 and 7 represent respectively the optimal con-trol and optimal value of the costThe curves start to increaseduring the first month because of the high infection level
6 ISRN Biomathematics
0 30 60 90 120 150 1800
05
1
15
2
25
3
Time (days)
times107
Rem
oved
R(t)
With controlWithout control
Figure 5 The function 119877 with and without control
0 30 60 90 120 150 1800
001
002
003
004
005
006
Time (days)
Con
trol
u(t)
Figure 6 The optimal control 119906lowast
0 30 60 90 120 150 1800
05
1
15
Time (days)
Cos
t
times10minus3
Figure 7 The optimal value of the cost
then they drop off steadily which is because of the constantand steady eradication of the infection
5 Conclusion
The work in this paper contributes to a growing literatureon applying optimal control techniques to epidemiology Weproposed amore realistic controlledmodel by including timedelay which represents the needed time for the migrationfrom the susceptible class to the recovered class after vac-cination The optimal control theory has been applied inthe context of an SIR model with time delay in state andcontrol variables and that includes a control 119906 that representsthe percentage of susceptible individuals being vaccinatedper time unit By using Pontryaginrsquos maximum principlethe explicit expression of the optimal controls was obtainedSimulation results indicate that the proposed control strategyis effective in reducing the number of susceptible and infectedindividuals and maximizing the recovered individuals
Acknowledgment
The authors would like to thank the anonymous referee forhisher valuable comments on the first version of the paperwhich have led to an improvement in this paper Researchreported in this paper was supported by the MoroccanSystemsTheory Network
References
[1] R M Anderson and R M May Infectious Diseases of HumansDynamics and Control Oxford University Press Oxford UK1992
[2] O Diekmann and J A P Heesterbeek Mathematical Epidemi-ology of Infectious Diseases John Wiley amp Sons Chisteter UK2000
[3] F Brauer and C C CastilloMathematical Models in PopulationBiology and Epidemiology Springer New York NY USA 2000
[4] P Van Den Driessche ldquoSome epidemiological models withdelaysrdquo inDifferential Equations andApplications to Biology andto Industry (Claremont CA 1994) pp 507ndash520 World Scien-tific River Edge NJ USA 1996
[5] Q J A Khan and E V Krishnan ldquoAn epidemic model with atime delay in transmissionrdquo Applications of Mathematics vol48 no 3 pp 193ndash203 2003
[6] N Yoshida and T Hara ldquoGlobal stability of a delayed SIRepidemic model with density dependent birth and death ratesrdquoJournal of Computational and AppliedMathematics vol 201 no2 pp 339ndash347 2007
[7] R Xu Z Ma and Z Wang ldquoGlobal stability of a delayed SIRSepidemic model with saturation incidence and temporaryimmunityrdquo Computers and Mathematics with Applications vol59 no 9 pp 3211ndash3221 2010
[8] XMeng L Chen and BWu ldquoA delay SIR epidemicmodel withpulse vaccination and incubation timesrdquo Nonlinear AnalysisReal World Applications vol 11 no 1 pp 88ndash98 2010
[9] G Zaman Y H Kang and I H Jung ldquoOptimal treatment of anSIR epidemic model with time delayrdquo BioSystems vol 98 no 1pp 43ndash50 2009
[10] K Hattaf and N Yousfi ldquoOptimal control of a delayed HIVinfection model with immune response using an efficientnumerical methodrdquo ISRN Biomathematics vol 2012 Article ID215124 7 pages 2012
ISRN Biomathematics 7
[11] G Birkhoff and G C Rota Ordinary Differential EquationsJohn Wiley amp Sons New YorkNY USA 4th edition 1989
[12] W H Fleming and R W Rishel Deterministic and StochasticOptimal Control Springer New York NY USA 1975
[13] D L Lukes Differential Equations Classical to Controlled Vol162 of Mathematics in Science and Engineering Academic PressNew York NY USA 1982
[14] L Gollmann D Kern and H Maurer ldquoOptimal control prob-lems with delays in state and control variables subject to mixedcontrol-state constraintsrdquo Optimal Control Applications andMethods vol 30 no 4 pp 341ndash365 2009
[15] K Hattaf and N Yousfi ldquoMathematical Model of the InfluenzaA(H1N1) Infectionrdquo Advanced Studies in Biology vol 1 no 8pp 383ndash390 2009
[16] M El hia O Balatif J Bouyaghroumni E Labriji and MRachik ldquoOptimal control applied to the spread of influenzaA(H1N1)rdquo Applied Mathematical Sciences vol 6 no 82 pp4057ndash4065 2012
[17] httpwwwpoumoncadiseases-maladiesa-zswineflu-grippe-porcinevaccines-vaccins fphp
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
ISRN Biomathematics 5
Step 1 For 119894 = minus119898 0 do 119878119894= 1198780 119868119894= 1198680 119877119894= 1198770 119906119894= 0
End forFor 119894 = 119899 119899 + 119898 do 120582119894
1= 0 120582119894
2= 0 120582119894
3= 0
End forStep 2 For 119894 = 0 119899 minus 1 do
119878119894+1=119878119894+ ℎ(Λ minus 119906
119894minus119898119878119894minus119898)
1 + ℎ(120573(119868119894119873) + 119889)
119868119894+1=
119868119894
1 + ℎ(120574 + 119889 + 120576 minus 120573(119878119894+1119873))
119877119894+1=119877119894+ ℎ(120574119868
119894+1+ 119906119894minus119898119878119894minus119898)
1 + ℎ119889
120582119899minus119894minus1
1=
120582119899minus119894
1+ ℎ(119860
1+ 120573(119868
119894+1119873)120582
119899minus119894
2+ (120582119899minus119894+119898
3minus 120582119899minus119894minus1+119898
1)120594[0119905119891minus120591]
(119905119899minus119894)119906119894)
1 + ℎ(119889 + 120573(119868119894+1119873))
120582119899minus119894minus1
2=120582119899minus119894
2+ ℎ(119860
2+ 120582119899minus119894
3120574 minus 120582119899minus119894minus1
1120573(119878119894+1119873))
1 + ℎ(120574 + 119889 + 120576 minus 120573(119878119894+1119873))
120582119899minus119894minus1
3=120582119899minus119894
3minus 1198603ℎ
1 + 119889ℎ
119879119894+1=(120582119899minus119894minus1+119898
1minus 120582119899minus119894minus1+119898
3)
1198604
120594[0119905119891minus120591]
(119905119894+1)119878119894+1
119906119894+1= min (119887max (0 119879119894+1))
Step 3 For 119894 = 0 n write119878lowast(119905119894) = 119878119894 119868lowast(119905119894) = 119868119894 119877lowast(119905
119894) = 119877119894 119906lowast(119905
119894) = 119906119894
End for
Algorithm 1
0
2
0 30 60 90 120 150 1800
5
Time (days)
times104 times10
6
With
out c
ontro
lI(t)
With
cont
rol
I(t)
Figure 2 The function 119868 with and without control
0 30 60 90 120 150 1800
05
1
15
2
Time (days)
times104
Infe
cted
I(t)
Figure 3 The function 119868 with control
Figures 2 3 and 4 represent the number of infectedindividuals (119868) with control (solid curve) and without control
0 30 60 90 120 150 1800
051
152
253
354
455
Time (days)
times106
Infe
cted
I(t)
Figure 4 The function 119868 without control
(dashed curve) It shows that in the presence of a controlthe number of infected individuals (119868) decreases greatly Themaximum number of infected individuals in the case withcontrol is 1787 times 104 and is 4404 times 106 in the case withoutcontrol then the efficiency of our strategy in reducing thespread of infection is nearly 9952
Figure 5 shows that the number of people removed beginsto grow notably from 24th day instead of 48th day inthe absence of control Moreover the number of recoveredindividuals at the end of the vaccination period is 2926 times107 instead of 2342 times 107 which represents an increase of
5840 times 106 cases
Figures 6 and 7 represent respectively the optimal con-trol and optimal value of the costThe curves start to increaseduring the first month because of the high infection level
6 ISRN Biomathematics
0 30 60 90 120 150 1800
05
1
15
2
25
3
Time (days)
times107
Rem
oved
R(t)
With controlWithout control
Figure 5 The function 119877 with and without control
0 30 60 90 120 150 1800
001
002
003
004
005
006
Time (days)
Con
trol
u(t)
Figure 6 The optimal control 119906lowast
0 30 60 90 120 150 1800
05
1
15
Time (days)
Cos
t
times10minus3
Figure 7 The optimal value of the cost
then they drop off steadily which is because of the constantand steady eradication of the infection
5 Conclusion
The work in this paper contributes to a growing literatureon applying optimal control techniques to epidemiology Weproposed amore realistic controlledmodel by including timedelay which represents the needed time for the migrationfrom the susceptible class to the recovered class after vac-cination The optimal control theory has been applied inthe context of an SIR model with time delay in state andcontrol variables and that includes a control 119906 that representsthe percentage of susceptible individuals being vaccinatedper time unit By using Pontryaginrsquos maximum principlethe explicit expression of the optimal controls was obtainedSimulation results indicate that the proposed control strategyis effective in reducing the number of susceptible and infectedindividuals and maximizing the recovered individuals
Acknowledgment
The authors would like to thank the anonymous referee forhisher valuable comments on the first version of the paperwhich have led to an improvement in this paper Researchreported in this paper was supported by the MoroccanSystemsTheory Network
References
[1] R M Anderson and R M May Infectious Diseases of HumansDynamics and Control Oxford University Press Oxford UK1992
[2] O Diekmann and J A P Heesterbeek Mathematical Epidemi-ology of Infectious Diseases John Wiley amp Sons Chisteter UK2000
[3] F Brauer and C C CastilloMathematical Models in PopulationBiology and Epidemiology Springer New York NY USA 2000
[4] P Van Den Driessche ldquoSome epidemiological models withdelaysrdquo inDifferential Equations andApplications to Biology andto Industry (Claremont CA 1994) pp 507ndash520 World Scien-tific River Edge NJ USA 1996
[5] Q J A Khan and E V Krishnan ldquoAn epidemic model with atime delay in transmissionrdquo Applications of Mathematics vol48 no 3 pp 193ndash203 2003
[6] N Yoshida and T Hara ldquoGlobal stability of a delayed SIRepidemic model with density dependent birth and death ratesrdquoJournal of Computational and AppliedMathematics vol 201 no2 pp 339ndash347 2007
[7] R Xu Z Ma and Z Wang ldquoGlobal stability of a delayed SIRSepidemic model with saturation incidence and temporaryimmunityrdquo Computers and Mathematics with Applications vol59 no 9 pp 3211ndash3221 2010
[8] XMeng L Chen and BWu ldquoA delay SIR epidemicmodel withpulse vaccination and incubation timesrdquo Nonlinear AnalysisReal World Applications vol 11 no 1 pp 88ndash98 2010
[9] G Zaman Y H Kang and I H Jung ldquoOptimal treatment of anSIR epidemic model with time delayrdquo BioSystems vol 98 no 1pp 43ndash50 2009
[10] K Hattaf and N Yousfi ldquoOptimal control of a delayed HIVinfection model with immune response using an efficientnumerical methodrdquo ISRN Biomathematics vol 2012 Article ID215124 7 pages 2012
ISRN Biomathematics 7
[11] G Birkhoff and G C Rota Ordinary Differential EquationsJohn Wiley amp Sons New YorkNY USA 4th edition 1989
[12] W H Fleming and R W Rishel Deterministic and StochasticOptimal Control Springer New York NY USA 1975
[13] D L Lukes Differential Equations Classical to Controlled Vol162 of Mathematics in Science and Engineering Academic PressNew York NY USA 1982
[14] L Gollmann D Kern and H Maurer ldquoOptimal control prob-lems with delays in state and control variables subject to mixedcontrol-state constraintsrdquo Optimal Control Applications andMethods vol 30 no 4 pp 341ndash365 2009
[15] K Hattaf and N Yousfi ldquoMathematical Model of the InfluenzaA(H1N1) Infectionrdquo Advanced Studies in Biology vol 1 no 8pp 383ndash390 2009
[16] M El hia O Balatif J Bouyaghroumni E Labriji and MRachik ldquoOptimal control applied to the spread of influenzaA(H1N1)rdquo Applied Mathematical Sciences vol 6 no 82 pp4057ndash4065 2012
[17] httpwwwpoumoncadiseases-maladiesa-zswineflu-grippe-porcinevaccines-vaccins fphp
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
6 ISRN Biomathematics
0 30 60 90 120 150 1800
05
1
15
2
25
3
Time (days)
times107
Rem
oved
R(t)
With controlWithout control
Figure 5 The function 119877 with and without control
0 30 60 90 120 150 1800
001
002
003
004
005
006
Time (days)
Con
trol
u(t)
Figure 6 The optimal control 119906lowast
0 30 60 90 120 150 1800
05
1
15
Time (days)
Cos
t
times10minus3
Figure 7 The optimal value of the cost
then they drop off steadily which is because of the constantand steady eradication of the infection
5 Conclusion
The work in this paper contributes to a growing literatureon applying optimal control techniques to epidemiology Weproposed amore realistic controlledmodel by including timedelay which represents the needed time for the migrationfrom the susceptible class to the recovered class after vac-cination The optimal control theory has been applied inthe context of an SIR model with time delay in state andcontrol variables and that includes a control 119906 that representsthe percentage of susceptible individuals being vaccinatedper time unit By using Pontryaginrsquos maximum principlethe explicit expression of the optimal controls was obtainedSimulation results indicate that the proposed control strategyis effective in reducing the number of susceptible and infectedindividuals and maximizing the recovered individuals
Acknowledgment
The authors would like to thank the anonymous referee forhisher valuable comments on the first version of the paperwhich have led to an improvement in this paper Researchreported in this paper was supported by the MoroccanSystemsTheory Network
References
[1] R M Anderson and R M May Infectious Diseases of HumansDynamics and Control Oxford University Press Oxford UK1992
[2] O Diekmann and J A P Heesterbeek Mathematical Epidemi-ology of Infectious Diseases John Wiley amp Sons Chisteter UK2000
[3] F Brauer and C C CastilloMathematical Models in PopulationBiology and Epidemiology Springer New York NY USA 2000
[4] P Van Den Driessche ldquoSome epidemiological models withdelaysrdquo inDifferential Equations andApplications to Biology andto Industry (Claremont CA 1994) pp 507ndash520 World Scien-tific River Edge NJ USA 1996
[5] Q J A Khan and E V Krishnan ldquoAn epidemic model with atime delay in transmissionrdquo Applications of Mathematics vol48 no 3 pp 193ndash203 2003
[6] N Yoshida and T Hara ldquoGlobal stability of a delayed SIRepidemic model with density dependent birth and death ratesrdquoJournal of Computational and AppliedMathematics vol 201 no2 pp 339ndash347 2007
[7] R Xu Z Ma and Z Wang ldquoGlobal stability of a delayed SIRSepidemic model with saturation incidence and temporaryimmunityrdquo Computers and Mathematics with Applications vol59 no 9 pp 3211ndash3221 2010
[8] XMeng L Chen and BWu ldquoA delay SIR epidemicmodel withpulse vaccination and incubation timesrdquo Nonlinear AnalysisReal World Applications vol 11 no 1 pp 88ndash98 2010
[9] G Zaman Y H Kang and I H Jung ldquoOptimal treatment of anSIR epidemic model with time delayrdquo BioSystems vol 98 no 1pp 43ndash50 2009
[10] K Hattaf and N Yousfi ldquoOptimal control of a delayed HIVinfection model with immune response using an efficientnumerical methodrdquo ISRN Biomathematics vol 2012 Article ID215124 7 pages 2012
ISRN Biomathematics 7
[11] G Birkhoff and G C Rota Ordinary Differential EquationsJohn Wiley amp Sons New YorkNY USA 4th edition 1989
[12] W H Fleming and R W Rishel Deterministic and StochasticOptimal Control Springer New York NY USA 1975
[13] D L Lukes Differential Equations Classical to Controlled Vol162 of Mathematics in Science and Engineering Academic PressNew York NY USA 1982
[14] L Gollmann D Kern and H Maurer ldquoOptimal control prob-lems with delays in state and control variables subject to mixedcontrol-state constraintsrdquo Optimal Control Applications andMethods vol 30 no 4 pp 341ndash365 2009
[15] K Hattaf and N Yousfi ldquoMathematical Model of the InfluenzaA(H1N1) Infectionrdquo Advanced Studies in Biology vol 1 no 8pp 383ndash390 2009
[16] M El hia O Balatif J Bouyaghroumni E Labriji and MRachik ldquoOptimal control applied to the spread of influenzaA(H1N1)rdquo Applied Mathematical Sciences vol 6 no 82 pp4057ndash4065 2012
[17] httpwwwpoumoncadiseases-maladiesa-zswineflu-grippe-porcinevaccines-vaccins fphp
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
ISRN Biomathematics 7
[11] G Birkhoff and G C Rota Ordinary Differential EquationsJohn Wiley amp Sons New YorkNY USA 4th edition 1989
[12] W H Fleming and R W Rishel Deterministic and StochasticOptimal Control Springer New York NY USA 1975
[13] D L Lukes Differential Equations Classical to Controlled Vol162 of Mathematics in Science and Engineering Academic PressNew York NY USA 1982
[14] L Gollmann D Kern and H Maurer ldquoOptimal control prob-lems with delays in state and control variables subject to mixedcontrol-state constraintsrdquo Optimal Control Applications andMethods vol 30 no 4 pp 341ndash365 2009
[15] K Hattaf and N Yousfi ldquoMathematical Model of the InfluenzaA(H1N1) Infectionrdquo Advanced Studies in Biology vol 1 no 8pp 383ndash390 2009
[16] M El hia O Balatif J Bouyaghroumni E Labriji and MRachik ldquoOptimal control applied to the spread of influenzaA(H1N1)rdquo Applied Mathematical Sciences vol 6 no 82 pp4057ndash4065 2012
[17] httpwwwpoumoncadiseases-maladiesa-zswineflu-grippe-porcinevaccines-vaccins fphp
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of