Upload
aitaoudia
View
213
Download
0
Embed Size (px)
Citation preview
7/28/2019 2009 VIZ-09 Edited Paper
1/6
2D Geometric Constraint Solving : an Overview
Samy Ait-Aoudia, Mehdi Bahriz, Lyes Salhi
ESI - Ecole nationale Suprieure en Informatique, BP 68M - Oued Smar 16270Algiers, Algeria
Abstract - Geometric constraint solving has applications in
many different fields, such as Computer-Aided Design,
molecular modelling, tolerance analysis, and geometric theorem
proving. Geometric modelling by constraints enables users to
describe shapes by relationships called constraints between
geometric elements. The aim is to derive automatically these
geometric elements and provide thus effort and time saving.
Moreover, users can easily modify existing designs. Many
resolution methods have been proposed for solving systems of
geometric constraints. We classify these methods in three broad
categories : algebraic, rule-oriented and graph-constructive
solvers.
Keywords - Constraints solving, geometric constraints,
algebraic, rule-oriented and graph-constructive solvers
I. INTRODUCTION
Geometric constraint solving has applications in manydifferent fields, such as Computer-Aided Design,
molecular modelling, tolerance analysis, and geometrictheorem proving. Geometric modelling by constraintsenables users to describe shapes by specifying a rough
sketch and adding to it geometric constraints i.e. a set ofrequired relations between geometric elements (see [6],
[27], [28]). The constraint solver must derive
automatically the correct shape needed. Typically, in 2D,geometric modelling by constraints specifies geometricalobjects such as points, lines, circles, conics by a set of
constraints : distances between points, points and lines,parallel lines, angles between lines, incidence relations
between points and lines, points and circles, tangency
relations between lines and circles or between circles, andso on.
Intuitively a dimensioned sketch is considered to be
well-constrained if it has a finite number of solutions for
non-degenerate configurations. Figure 1 shows a 2Dwell-constrained problem with three distance constraints
between pairs of points and two angle constrains.Similarly a dimensioned sketch is considered to be under-
constrained if it has an infinite number of solutions for
non-degenerate configurations. Finally a dimensionedsketch is considered to be over-constrained if it has no
solutions for non-degenerate configurations. Examples of
2D under-constrained and over-constrained sketches aregiven in figures 2.a and 2.b respectively.
B
D
d2
d3
C
d1
Figure 1. A simple constrained design.
(a) (b)
Figure 2. (a) an under-constrained sketch, (b) an over-constrained
sketch
Many resolution methods have been proposed forsolving systems of geometric constraints. They can be
classified in many ways. We classify these methods in
three broad categories : algebraic, rule-constructive andgraph-constructive solvers.
This paper is organized as follows. The algebraic
approach for constraints solving is explained in section II.We present in section III, the rule-oriented methods to
handle constraint problems. The graph-based
methodology is given in section IV. Section V givesconclusions.
II.ALGEBRAIC APPROACH
A.Numerical methodsIn a numerical constraint solver, the geometric
constraints are first translated into a system of algebraicequations (linear or not). This system is then solved by
applying a numerical method. For each constraint
correspond an algebraic equation. As an example of thiscorrespondence, the equation corresponding to "internal"
2009 Second International Conference in Visualisation
978-0-7695-3734-4/09 $25.00 2009 IEEE
DOI 10.1109/VIZ.2009.29
201
2009 Second International Conference in Visualisation
978-0-7695-3734-4/09 $25.00 2009 IEEE
DOI 10.1109/VIZ.2009.29
201
2009 Second International Conference in Visualisation
978-0-7695-3734-4/09 $25.00 2009 IEEE
DOI 10.1109/VIZ.2009.29
201
2009 Second International Conference in Visualisation
978-0-7695-3734-4/09 $25.00 2009 IEEE
DOI 10.1109/VIZ.2009.29
201
7/28/2019 2009 VIZ-09 Edited Paper
2/6
and "external" tangency between two circles (see figure
3), one centred atA with radius r1, the second centred at
B with radius r2 is the following :
(xB - xA)2 + (yB - yA)
2 - (r1 r2)2= 0.
r2r1
A B
C1
C2
B
r2
C1
C2
r1
A
Figure 3. Tangency between two circles.
Example :
A constrained scheme is shown in figure 4. It consists
of three points and three distance constraints. The point
P1 lies at the origin and P3 on the positive x-axis. The
corresponding set of equations is given below :
eq1 : x1 = 0.eq
2: y
1= 0.
eq3
: y3
= 0.
eq4
: y2
y1
d1
= 0.
eq5
: (x3 x
2)
2
+ (y3
y2)
2
d3
2
= 0.
eq6
: x3
x1
d2
= 0.
Sutherland [55], Hillyard and Braid [24], Borning [7]
Borning and Duisberg [8] and Gosling [22] have used therelaxation method to solve the constrained problem.
Relaxation is a quite slow method and sometimes
convergence can't be achieved. Barford [5] used a
projective method.
Many systems have used the Newton-Raphsonsiteration to solve the set of geometric constraints. This
"popular" numerical method was used, among many
others, by Lin et al. [41], Light et al. [40], Lee et al. [37],
Nelson [45], Prusinkiewicz et al. [49], Rocheleau et al.
[50], Gossard et al. [23], Serrano [51], Perez et al [47],
Anderl et al. [4]. This method needs an initial guess,
typically given by the sketch of the desired geometric
scheme.
(x2,y2)
(x3,y3)(x1,y1)
d1
d2
d3
x
y
P1 P3
P2
Figure 4. A constraint problem.
However, there is a well-known problem. If Newton-
Raphsons method often works well, sometimes it does
not converge or it converges to an unwanted solution. Inthis last case, the user changes his initial guess until
Newton-Raphsons method works if it does.
An alternative method to Newton-Raphson for
geometric constraint solving is homotopy or continuation
[3]. Lamure et al. [34] have tested several configurationsusing homotopy with success and where Newton-Raphson's method fails. Nonetheless, homotopy, is
slower than Newton's method. Homotopy was also used
for solving 3D constraints by Durand et al. [14]. Foufou
et al. [16] recently used efficient numerical methods to
solve systems of constraints.
B.Symbolic methodsAs in the numerical solvers, the constraints are again
formulated as a system of algebraic equations. However,
instead of applying numerical techniques to determine a
solution, general symbolic computations are undertaken
to find the solution to the system of equations. Methods
such as Grbner basis or Wu-Ritt [32] techniques can be
applied to find symbolic expressions for the solutions.
Many systems have used a symbolic resolution of the
system of equations. We can non exhaustively mention
works done by Ericson et al. [15], Kondo [31], Buchanan
et al. [12], Chou et al. [13] and Gao et al [18].
III. RULE-CONSTRUCTIVE APPROACH
A.Problem formulationRule-based solvers rely on the predicates
formulation. The constraints are expressed as facts and aninference engine is used to derive the solution byexhaustively applying rules.
Some examples of expressing constraints as facts aregiven below :
coord(P1, [X, Y])coordinates of point P1 are X and Y.
dist(P1, P2, d)d is the distance between points P1 et P2.
slope(P1, P2, a)a is the slope of the line P1P2.
angle(P1, P2, P2, b)b is the angle P1P2P3.
slope(P1, P2) = slope(P3, P4)line P1P2 is parallel to line P3P4.
A typical example of rule used by the inferenceengine is given hereafter :
[coord(P1, [X1, Y1]), coord (P2, [X2, Y2]), dist(P1, P3, R1),dist(P2, P3, R2)] coord(P3, intersection(circle(c(P1), R1), circle(c(P2),
R2)))]
202202202202
7/28/2019 2009 VIZ-09 Edited Paper
3/6
It states that if points P1and P2are known and thedistance values R1 and R2between points (P1,P3) and(P2,P3) respectively are given then we can compute thecoordinates of the point P3 by intersecting two circles,one centered atP1with radius d13 the other centered atP2with radius d23.
B.Rule-Oriented solversWe will review now some works done in the rule-
oriented field. The majority of rule-constructive solvershave been implemented at the end of the eighties andearly nineties.
Brderlin ([10], [11]), has developed a system takinginto account 2D and 3D constraints. This system wasimplemented using Prolog. The class of 2Dconfigurations solved was not very large and the 3D casesolves very simple schemes.
Sunde in [53] uses a rule-constructive method. Heintroduces the notion of CA (Constraint Angle) and CD(Constraint Distance) sets. The method tries to merge allthe initial CA and CD sets in a single CD set. If themerging operation succeeds the figure solution isconstructed.
Aldefeld in [2] uses a forward chaining inferencemechanism, where the notion of direction of lines isimposed by introducing additional rules, and thusrestricting the solution space.
A similar method was presented by Suzuki et al. [54],where handling of over-constrained and under-constrained problems is given special consideration.
Verroust et al [56] have used the same theoreticalfoundations as Sunde [53]. They added some rules toextent the class of problems solved.
Rule-constructive approach provide a qualitativestudy of geometric constraints. Although it is a goodapproach for prototyping and experimentation, theextensive computations involved in the exhaustivesearching and matching make it inappropriate for real
world applications.IV. GRAPH-BASED APPROACH
A.Graph representationIn graph-based methods, an undirected graph
G=(V,E) where |V|=n and |E|=m represents the
constraint problem. The geometric elements are
represented by the graph nodes and the constraints are the
graph edges. The class of configurations solved by these
methods is typically the ruler and compass constructive
problems. In Computer-Aided Design, the graph-based
approach has become dominant.
Example:
A dimensioning scheme defining a constraint problemis shown in figure 5. It involves six points and six lines.
The constraints are six point-point distances, three line-
line angles, twelve point-line implicit distances that are
zero. The corresponding constraint graph is shown in
figure 6. The edges are labeled with the values of the
distance and angle dimensions. The unlabeled edges
correspond to the implicit point-line distances.
A
B
C
D
E
F
d1 d2
d3
d4d5
d6
Figure 5. A constraint problem.
A
B
C
D
E
F
AB BC
CD
DEEF
AF
d1 d2
d3
d4d5
d6
Figure 6. The constraint graph.
B.Graph structure analysisGraph-constructive solvers are stemming from graph
theory. They are based on an analysis of the structure of
the constraint graph. The graph constructive approachprovides means for developing sound and efficient
algorithms. Let us now give some definitions concerning
the structural properties of the constraint graph when the
degree of freedom of each geometric object is two and
each constraint 'fixes' one degree of freedom.
Definition 1
A constraint graph G=(V,E) where |V|=n and |E|=m
is structurally well-constrained if and only ifm=2*n-3
and m2*n-3 for any induced sub-graph G=(V,E)where |V|=nand |E|=m(see [33]).
Definition 2
A constraint graph G=(V,E) contains a structurallyover-constrained part if there is an induced sub-graph
G=(V,E) having more than 2*n-3 edges.
Definition 3
A constraint graph G=(V,E) is structurally under-
constrained if it is not over-constrained and the
number of edges is less than 2*n-3.
203203203203
7/28/2019 2009 VIZ-09 Edited Paper
4/6
We must mention that the Laman's theorem [33] is
proved only for point to point distance constraints and
that the extension of his definitions may imply incorrect
cases. A typical example of such cases is given by a
triangle constrained with three angle constrains. In figure
7.left, a triangle is constrained with three angle
constraints (with ++=180). The triangle is
geometrically under-constrained but the constraint graphrelated to (figure 7.right) is well constrained.
Figure 7. An under-constrained problem (left), and its associated
constraint graph (right).
The graph constructive approach uses only thestructure of the constraint graph and forgets numerical
information. A constraint graph can be structurally well-
constrained but numerically under-constrained. The
dimensioned scheme shown in figure 8 (left) is
numerically under-constrained (the edges are labeled with
distance arguments). Its corresponding constraint graph
shown in figure 8 (right) is structurally well-constrained.
Nonetheless, with the graph constructive approach, such
cases will be detected during the construction phase.
A B
C
1
0
D
1
1
1
D
A B
C
Figure 8. A numerical under-constrained problem (left), and its
associated constraint graph (right).
C.Resolution principleGraph-based algorithms for solving geometric
constraint problems have two phases : an analysis phase
and a construction phase. These algorithms are also called
decomposition recombination methods.
During the first phase the graph of constraints is
analyzed and decomposed in small sub-problems.
Sequences of construction steps are derived. During the
second phase the recombination is carried out and theconstruction steps are processed to place the geometric
elements. Figure 9 shows an example of decomposition
of a constrained scheme.
The placement steps correspond to standardized
geometric construction steps (place a point at given
distances from two points, place a line at prescribed angle
from another line through a point, rotate and translate a
sub-structure, ).
Figure 9. Constrained Design decomposition.
Figure 10 shows an example of a placement step. The
point p3 is to be constructed from two known points p1
andp2. dij is the value of the constraint distance between
points pi an pj. To obtain point p3, we intersect two
circles, one centered at p1 with radius d13 the other
centered at p2 with radius d23. If the point to pointdistance constraint between p1 and p2 is the first
constraint "pinned", we assume that p1 is placed at the
origin and p2on the positive x-axis at distance d12 from
p1.
These approaches are fast and more methodical. In
this methods the constraints are satisfied in a constructive
manner, which makes the constraint solving process
natural for the user and suitable for interactive debugging.
Figure 10. Placing a point relative to two known points.
D.Graph based solversIn the early nineties, Owen [46] was the first to give
an efficient and sound graph-based algorithm. This
algorithm is quadratic and performs well on ruler and
compass constructive problems. Following that, based on
Owen's work, was developed a commercial systemnamed D-Cubed. Since that and until recently, a lot of
researches were conducted in this promising way. Among
these works we can mention those done by Hoffmann's
"constraint team" ([9], [17], [42], [43], [44], [25], [26],
[19]), Lamure et al. [35], Lee et al. [38], Ait-Aoudia et al.
[1], Jermann et al. [29], Gao et al [20], Lee et al. [39],
Sitharam et al [52], Gao et al. [21], Zhang et al. [57],
Joan-Arinyo et al. [30], Podgorelec et al. [48].
xp1 p2
p3
p3
d12
d13 d23
y
A
B C
AB
AC BC
AB
C
1
2
3
6
5
7
4
9
8
10
11
12
13
14
16
15
204204204204
7/28/2019 2009 VIZ-09 Edited Paper
5/6
V.CONCLUSION
We have classified the geometric constraint solvers inthree broad categories. we summarise hereafter someadvantages and drawbacks for each approach.
Numerical methods are O(n3) or worse. Thesemethods suffer from the lack of "geometric explanation"during the resolution process. Also, most numerical
methods have difficulties for handling over-constrainedor under-constrained schemes. The advantage of thesemethods is that they have the potential to solve largenonlinear system that may not be solvable using any ofthe other methods. Almost all existing solvers switch toalgebraic methods when the given configuration is notsolvable by the native method. Algebraic methods handleeasily the 3D case. Speeding up the resolution must beconsidered further.
Symbolic methods are typically exponential in time
and space. They can be used only for small systems.
According to Lazard [36], computing the Grbner bases
of an irreducible system of degree two in ten unknowns is
a hopeless case.
Rule-based solvers rely on the predicates formulation.
Although they provide a qualitative study of geometricconstraints, the "huge" amount of computations needed
(exhaustive searching and matching) make them
inappropriate for real world applications.Graph-based methods can be very efficient. In
Computer-Aided Design, the graph-based approach hasbecome dominant. However, these methods are onlyapplicable to particular kinds of problems, typically rulerand compass constructive problems. The graph-basedapproach to handle efficiently the 3D case deservesfurther studies. The theoretical framework is no longerapplicable. The extension of Laman theory (sectionIV.B)to the 3D case will give: a constraint graph G=(V,E)where |V|=n and |E|=m is structurally well-constrained ifand only if m=3*n-6and m3*n-6 for any induced
sub-graph G=(V,E) where |V|=n and |E|=m. The3D constrained design shown in figure 11 involves eight
points and eighteen point to point distances. This wellknown example verifies the formula given above but it isnot structurally well constrained. The two sub-partsdefined by the points {A,B,C,D,E} and {A,G,H,F,E}canrotate around theAEaxis.
Figure 11. A 3D constrained design.
VI. REFERENCES
[1] S. Ait-Aoudia, H. Brahim, A. Moussaoui , T. Saadi.Solving Geometric Constraints by a Graph-Constructive
Approach. IEEE International Conf. on InformationVisualization , (London, England), pp. 250-255, July 99.
[2] B. Aldefeld. Variation of geometries based on a geometricreasoning method. CAD vol.20, num.3, April 88. pp 117-
126.[3] E.L. Allgower and K. Georg. Continuation and path
following. Acta Numerica, pages 1-64, 1993.[4] R. Anderl and R. Mendgen. Parametric design and its
impact on solid modeling applications. Symposium onSolid Modeling and CAD/CAM Applic. May 95. pp. 1-12.
[5] L.A. Barford. A graphical language-based editor forgeneric solid models represented by constraints. PhDThesis, Cornell University, May 1987.
[6] B. Bettig, J. Shah. Derivation of a standard set ofgeometric constraints for parametric modeling and dataexchange. CAD, 33 (2001), 17-33.
[7] A. Borning. THINGLAB: a constraint-oriented simulationlaboratory. Ph.D Thesis, Stanford University 1979.
[8] A. Borning and R. Duisberg. Constraint-based tools forbuilding user interfaces. ACM Transactions on Graphics,vol. 5, num. 4, Oct. 86. pp 345-374.
[9] W. Bouma, I. Fudos, C.M. Hoffman, J. Cai, R. Paige. Ageometric constraint solver. Computer Aided Design, Vol.27, No.6, June 1995, 487-501.
[10]B. Brderlin. Constructing three-dimensional geometric
objects defined by constraints. Proceedings 1986.Workshop on Interactive 3D Graphics. pp.111-129, ChapelHill, N.C, Oct 86.
[11]B. Brderlin. Automatizing geometric proofs andconstructions. Computational Geometry and itsApplications. pp 233-252. Lecture notes in ComputerScience, 333, Springer Verlag, Mars 88.
[12]S.A. Buchanan, A. Pennington. Constraint definition
system : a computer-algebra based approach to solvinggeometric constraint problems. CAD, vol. 25, num. 12,Dec.93. pp 741-750.
[13]S.C. Chou, W.F. Schelter and J.G. Yang. Characteristic
sets and Grbner bases in geometry theorem proving.Workshop on Computer-Aided Geometric Reasoning,INRIA, Sophia Antipolis, June 87. pp 29-56.
[14]C. Durand, C.M. Hoffman. Variational Constraints in 3D.Proc. Intl Conf on Shape Modeling and Appl, Aizu, Japan,1999; 90-97.
[15]L.W. Ericson, C.K. Yap. The design of LINETOOL ageometric editor. Fourth Annual Symposium onComputational Geometry, ACM, June 88, pp 83-92.
[16]S. Foufou, D. Michelucci, J.P. Jurzak. NumericalDecomposition of Geometric Constraints, Proceedings ofthe 2005 ACM symposium on Solid and physicalmodeling, Cambridge, Massachusetts, pp. 143 - 151.
[17]Fudos, C.M. Hoffman. A Graph-Constructive approach toSolving Systems of Geometric Constraints. ACM Trans.on Graphics, Vol.16, N2, April 1997, 179-216.
[18]X.S. Gao and S.C. Chou. Solving geometric constraintsystems : a symbolic approach and decision of Rc-constructibility. Computer Aided Design, pp. 115-122. vol.30, n. 3, 1998.
[19]X.S. Gao , C.M. Hoffman and W. Yang. Solving Spatial
Basic Geometric Constraint Configurations with LocusIntersection. Solid Modeling '02, 2002.
[20]X.S. Gao and G.F. Zhang. Geometric Constraint SolvingBased on Connectivity of Graph. MMRC, AMSS,
Academia Sinica, Beijing, n22, Dec2003, pp148-162.
A
E
B C
D
G H
F
205205205205
7/28/2019 2009 VIZ-09 Edited Paper
6/6
[21].S. Gao, Q. Lin, G.F. Zhang. A C-tree decomposition
algorithm for 2D and 3D geometric constraint solving,
Computer-Aided Design, Vol. 38, Issue 1, January 2006,
Pages 1-13.
[22]J. Gosling. Algebraic constraints. Ph.D Thesis, Carnegie-
Mellon University, May 83.
[23]D. Gossard, R. Zuffante and H. Sukurai. Representing
dimensions, tolerances and features in MCAE systems.IEEE Computer Graphics and Applications, Mars 88. pp
51-59.
[24]R. Hillyard and I. Braid. Characterizing non ideal shapes in
terms of dimensions and tolerances. Computer Graphics,
vol. 12, num. 3, Aug. 78. pp 234-238.
[25]C.M. Hoffman and C.S. Chiang. Variable-Radius Circles
of Cluster Merging in geometric constraints Part I:
Translational Clusters. Computer Aided Design, vol. 33,
2001.
[26]C.M. Hoffman and C.S. Chiang. Variable-Radius Circles
in Cluster Merging : Part II: Rotational Clusters. Computer
Aided Design, vol. 33, 2001.
[27]C.M. Hoffman and R. Joan-Arinyo. A Brief on Constraint
Solving, Computer-Aided Design and Applications,
2(5):655-663, 2005..[28]C.M. Hoffman. Summary of basic 2D constraint solving,
Int. Journal Product Lifecycle Management, Vol. 1, No. 2,
2006.
[29]C. Jermann, G. Trombettoni, B. Neveu, M. Rueher. A
constraint programming approach for solving rigid
geometric systems. 6th Conf. on Principles and Practice of
constraint programming pp 233-248, Singapore sep. 2000.
[30]R. Joan-Arinyo, M. Tarrs-Puertas, and S. Vila-Marta.
Geometric Constraint Graphs Decomposition Based on
Computing Graph Circuits, Seventh International
Workshop on Automated Deduction in Geometry
September 22-24, 2008
[31]K. Kondo. Algebraic method for manipulation of
dimensional relationships in geometric models. CAD
vol.24, num.3, March 92. pp 141-147.[32]B. Kutzler. Introduction to Bucheberger's Grbner bases
method. Workshop on Computer-Aided Geometric
Reasoning, INRIA, Sophia Antipolis, June 87, pp 7-25.
[33]G. Laman. On graphs and rigidity of plane skeletal
structures. Journal of Engineering Mathematics, vol.4,
num.4, Oct. 70. pp 331-340.
[34]H. Lamure and D. Michelucci. Solving constraints by
homotopy. Symposium on Solid Modeling Foundations
and CAD/CAM Applications. May 95. pp. 263-269.
[35]H. Lamure and D. Michelucci. Qualitative study of
geometric constraints. Geometric Constraint Solving and
applications, B. Brderlin and D. Roller editors, Springer-
Verlag, 1998.
[36]D. Lazard. Systems of algebraic equations : algorithms and
complexity. Rapport interne LITP 92.20, Mars 92.
[37]K. Lee and G. Andrews. Inference of the positions of
components in an assembly : part 2. CAD vol.17, num.1,
Jan. 85. pp 20-24.
[38]J.Y. Lee, K. Kim. A 2D geometric constraint solver using
DOF-based graph reduction. Computer Aided Design. Vol.
30, No. 11, 1998, pp. 883-896.
[39]K.Y. Lee, O.H. Kwon, J.Y. Lee and T.W. Kim. A hybrid
approach to geometric constraint solving with graph
analysis and reduction. Advances in Engineering Software,
34 (2003), pp 103-113.
[40]R. Light and D. Gossard. Modification of geometric
models through variational geometry. CAD vol.14, num.4,
July 82. pp 209-214.
[41]V. Lin, D. Gossard and R. Light. Variational geometry in
CAD. Computer Graphics, vol.15, num.3, Aug.81. pp 171-
178.[42]A. Lomonosov ,M. Sitharam and C.M. Hoffman. Finding
Solvable Subsets of Constraint Graphs. Springer LNCS
1330, G. Smolka, ed., 1997, 463-477.
[43]A. Lomonosov ,M. Sitharam and C.M. Hoffman.
Geometric Constraint Decomposition. Geometric Constr
Solving and Appl., B. Bruderlin and D. Roller, eds, 1998,
Springer, 170-195.
[44]A. Lomonosov ,M. Sitharam and C.M. Hoffman
Decomposition Plans for Geometric Constraint Problems,
Part II: New Algorithms. Journal of Symbolic
Computations 31, 2001, 409-428.
[45]G. Nelson. Juno, a constraint-based graphics system.
Computer Graphics, vol. 19, num. 3, 1985. pp 235-243.
[46]J.C. Owen. Algebraic solution for geometry from
dimensional constraints. Proceeding Symposium on SolidModeling Foundations and CAD/CAM Applications,
Austin Texas. June 91. pp. 397-407.
[47]A. Perez and D. Serrano. Constraint base analysis tools for
design. Second Symposium on Solid Modeling
Foundations and CAD/CAM Applications, Montreal
Canada. Mai 93. pp. 281-290.
[48]D. Podgorelec, B. Zalik, V. Domiter. Dealing with
redundancy and inconsistency in constructive geometric
constraint solving, Advances in Engineering Software 39
(2008) 770786.
[49]P. Prusinkiewicz and D. Streibel. Constraint-based
modeling of three-dimensional shapes. Graphics Interface
'86, Vision Interface '86, 1986. pp 158-163.
[50]D.N. Rocheleau and K. Lee. System for interactive
assembly modelling. CAD vol. 19, num.2, Mars 87. pp 65-72.
[51]D. Serrano. Automatic dimensioning in design for
manufacturing. Proceeding Symposium on Solid Modeling
Foundations and CAD/CAM Applications, Austin Texas.
June 91. pp. 379-386.
[52]M. Sitharam, J.J. Oung, Y. Zhou and A. Arbree. Geometric
constraints within Feature Hierarchies, Computer-Aided
Design, Volume 38, Issue 1, January 2006, Pages 22-38.
[53]H. Suzuki, H. Ando and F.Kimura. Geometric constraints
and reasoning for geometrical CAD systems. Computer
and Graphics, vol. 14, num. 2, Oct. 90. pp 211-224.
[54]G. Sunde. Specification of shape by dimensions and other
geometric constraints. Geometric modeling for CAD
applications, pp. 199-213. North-Holland, IFIP, 1988.
[55]I.E. Sutherland. SKETCHPAD: a man machine graphical
communication system. Ph.D Thesis, MIT, Mass. 1963.
[56]A. Verroust, F. Schonek and D. Roller. Rule oriented
method for paremetrized computer-aided design. CAD,
vol. 24, num. 3, Oct 92. pp 531-540.
[57]G.F. Zhang, X.S. Gao. Spatial geometric constraint solving
based on k-connected graph decomposition, 2006 ACM
symposium on Applied computing, Dijon France, pp. 979-
983.
206206206206