Upload
vuongkhanh
View
222
Download
2
Embed Size (px)
Citation preview
©20
06 T
he M
athW
orks
, Inc
.
Terry Denery, Ph.D.
Physics-Based Modeling Tools
Modeling Mechanical, Electric, and Hydraulic Systems in Simulink®
2
Physics-Based Modeling Methods Improve Control System Design
Multidomain systems (mechanical, electrical, hydraulic, chemical, . . .)Successful controller development requires thorough and accurate understanding of plant
ControllerElectricalMechanicalDevice
Actuators
Sensors
Plant+u y
3
ElectricalMechanicalDevice
Actuators
Sensors
Plant
Important Physical Domains
Mechanical-3D Multi-Body Dynamics-Driveline Mechanics SimDriveline
SimMechanics
4
ElectricalMechanicalDevice
Actuators
Sensors
Plant
Important Physical Domains
Electrical-Circuits-Motors and Actuators-Power Systems
SimPowerSystems
5
ElectricalMechanicalDevice
Actuators
Sensors
Plant
Important Physical Domains
Hydraulics-Circuits-Motors and Actuators-Power Systems
SimHydraulics
9
Electrical Power NetworkHydraulic Power Network
Mechanical Driveline Power Network
Electrical
Mechanical
Device
10
Plant
Electrical
Mechanical
Device
Actuators
SensorsController+u y
Control Systems within Control Systems
11
Biggest Benefit – Delivered Through Simulink®
Rich Modeling Environment– Physical– Behavioral– Data-Driven
Control System Development Tool– One environment for controller and plant– Code generation that enables HIL testing– Easy access to control tools
12
Why did we build SimHydraulics?
Hydraulics is widely used in motion actuation systemsCustomer Feedback – “Extend benefits of Model-Based Design to hydraulic
system design”– “Develop better plant models for better controller
development”– “Co simulation with other products is difficult and
impractical”
13
Pros and Cons for Hydraulics
Pros:High power densityStraight-line actuators are simple, efficient and reliableVariable speed is easily obtainedSimple, efficient, and centralized control Overload protection
Cons:High costComplex and costly maintenanceExtremely vulnerable to dirt and contaminationPossible leaksNoise and vibrationsFire hazards
Design Trend is Electro-Hydraulics:Transmit power electrically, but deliver through local hydraulic networks
15
Block Diagrams
( ) ( ).,,
,
,,
),(
211
1211
TAAPTP
OOAOP
AA
TAOAPO
pp
ppppppqqqqq
dtdpKqA
ppKqppKq
tpp
−+−=−−==
⋅=
−=−=
=
21
First-Principles Modeling
Use an understanding of the system’s physics to derive a mathematical representation
∫ −−−−−−−−−−+−
=)(sin1
)cos()cos())sin(()sin())sin(()sin(2
2222
γαγγααγαγαγγααα
nennenwL &&
& dγ
Two Approaches to Modeling Dynamic Systems
22
First-Principles Modeling Data-Driven Modeling
Use system test data to derive a mathematical representation
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
+++
+++
=
13
231
)(
2
2
23
sss
sss
sH
Use an understanding of the system’s physics to derive a mathematical representation
∫ −−−−−−−−−−+−
=)(sin1
)cos()cos())sin(()sin())sin(()sin(2
2222
γαγγααγαγαγγααα
nennenwL &&
& dγ
Two Approaches to Modeling Dynamic Systems
23
First-Principles Modeling Data-Driven Modeling
Advantages:Provides insight into the system’s underlying behaviorEnables performance prediction for unbuilt systems
Disadvantages:Effects like friction and turbulence are difficult to characterizeMay be time consuming to develop
Advantages:Can be a fast method for developing an accurate modelInstills confidence because it uses data from an actual system
Disadvantages:Requires a physical system to acquire test dataLacks description of physics of the systemMay need multiple data sets to cover range of system operation
Each Approach Has Its Merits
24
Simulink®
ParameterEstimation
Complete Modeling Environment
Simulink
SimHydraulicsSimMechanicsSimDriveline
SimPowerSystems
First-Principles
Neural NetworkToolbox
SystemIdentification
Toolbox
Fuzzy LogicToolbox
Data-Driven
Tools that Span Both Modeling Approaches
25
Inform Model with Test Data
Data fitting and optimization techniquesSets parameters defined by physics-based modeling tools
Simulink Parameter Estimation
26
Code Generation with Real-Time Workshop®
Generate ANSI C codeHardware-in-the-loop (HIL) simulationsS-functions In-house code
27
MATH EngineeringSimulation
MATLAB &Simulink
DomainKnowledge
ProgrammingEffort
Flexibility of mathematical modeling enables you to simulate almost anything.
General Engineering Simulation
28
MATH HydraulicSimulation
SimHydraulics Dom.Know.
Prog.Effort
SimHydraulics brings you further by reducing your requirementsfor hydraulic domain knowledge and programming effort.
MATLAB &Simulink
DomainKnowledge
ProgrammingEffort
Hydraulic Engineering Simulation
29
MATH MultidomainSimulation
The MathWorks provides the best combination of tools for multidomain simulations.
Multidomain Simulation
Controls Tools DK PE
Real Time Workshop DK. PE
Parameter Estimation DK. PE
SimMechanics DK. PE
MATLAB &Simulink DK PE
SimPowerSystems DK. PE
SimDriveline DK. PE
SimHydraulics DK. PE