Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
1© 2019 The MathWorks, Inc.
Mallipohjaisen suunittelun ja koneoppimisen
hyödyntäminen ennakoivassa huollossa
Markus Orpana
Application Engineer
2
Key Takeaways
1. Data Analytics in Industrial Automation and Machinery
2. Integrated Matlab/Simulink Platform
3
Increasing amount of Data
Increasing amount of
▪ Business Data & Engineering Data
▪ Cheaper sensors, communication
and data storage
▪ Powerful algorithms (e.g. machine
learning, deep learning)
▪ Applications: Machine health
monitoring and predictive
maintenance
4
ChallengeReduce waste and machine downtime in plastics
manufacturing plants
SolutionUse MATLAB to develop and deploy monitoring and
predictive maintenance software that uses
machine learning algorithms to predict machine
failures
Results▪ More than 50,000 euros saved per year
▪ Prototype completed in six months
▪ Production software run 24/7
Mondi Implements Statistics-Based Health
Monitoring and Predictive Maintenance for
Manufacturing Processes with Machine
Learning
Link to user story
“MathWorks Consulting’s support
is among the best I’ve seen; the
consultants are fast and
exceptionally knowledgeable.
We’ve already seen a positive
return on investment from cost
savings, and now we have more
budget and time to complete
more machine learning projects
that will provide similar benefits.”
Dr. Michael Kohlert
Mondi
One of Mondi Gronau’s plastic production
machines, which deliver about 18 million tons of
plastic and thin film products annually.
Link to video
5
What is Predictive Maintenance?
6
Pump - detected
I need help.
7
Pump - detected
I need help. One of my
cylinders is blocked. I will
shut down your line in 15
hours
8
Fault Classification Algorithms Allow You to Identify the Root
Cause of Anomalous Behavior
▪ Three-phase pump commonly used
for drilling and servicing oil wells
– Three plungers try to ensure a uniform
flow
▪ Condition monitoring to detect:
– Seal leak
– Inlet blockage
– Bearing degradation
Component
Failure
Crankshaft
Outlet
Inlet
9
Fault Classification Algorithms Allow You to Identify the Root
Cause of Anomalous Behavior
Component
Failure
Crankshaft
Outlet
Pressure & FlowSensor
Inlet
▪ Three-phase pump commonly used
for drilling and servicing oil wells
– Three plungers try to ensure a uniform
flow
▪ Condition monitoring to detect:
– Seal leak
– Inlet blockage
– Bearing degradation
▪ Identify fault present in system using
only pressure and flow sensor data
10
Generate Synthetic Failure Data from Simulink Models if Real
Failure Data is Unavailable
▪ Model failure modes
– Work with domain experts and the data
available
– Vary model parameters or components Simulink Model
Build
model
Inject Failures
Incorporate
failure modes
Generated
Failure Data
Run
simulations
11
Workflow -- Predictive Maintenance with Synthetic Data
1. Modelling
2. Data Generation and Pre-processing
3. Feature Extraction
4. Classification
12
Simscape Pump Model
13
Feature Extraction - Obtaining Distinctive Features
Time Domain Features
Frequency Domain Features
14
Feature Extraction
15
Classification
17
Condition
Monitoring
What have we achieved?
Why is my
machine behaving
abnormally?
Anomaly
Detection
Is my machine
operating
normally?
18
Key Takeaways
1. Using simulation model/digital twin for generating synthetic failure data
2. Interactive tools for maintenance/machine learning algorithm development
19
Kiitos