MATLAB, Simulink, Simscape, SimPowerSystems ...

MATLAB, Simulink, Simscape, SimPowerSystems, xPC Target: Modelización y prototipado de sistemas eléctricos y electrónicos de potencia

Hotel ME, Madrid 2 octubre 2012

© 2012 The MathWorks, Inc.1

MathWorks Vital Statistics Developers of MATLAB & Simulink 2,400 staff worldwide

Support staff worldwide Development staff in Natick, MA 30% of revenue invested in R&D $700M annual revenue 2012 - orders from 23,000 companies in 128 countries

22

Key Capabilities Drive MathWorks Business

Verification, Validation, and Test

•

Automatic Code Generation

Test and measurement

•

Embedded software

•

•

• •

Model checking

•

DSP software

Code verification Qualification kits

VHDL/Verilog

•

PLC code

• Discrete-event modeling  Simulink • DSP designs • State charts • Physical modeling

System Modeling and Simulation

• •

Data Analysis and Algorithm Development

Technical Computing

Rapid prototyping and HIL

•

Control design Signal processing

• •

Optimization Statistics

•

Communication systems

•

Image processing

• Application

 MATLAB

deployment

•

Computational finance

• Student version • Instrument and database

•

Video processing

•

Computational biology

• Distributed and

parallel computing

connectivity

1985

1990

1995

2000

2005

• MATLAB

Mobile for iPhone 2010

Founded in 1984

3

MathWorks Investment in Physical Modeling

Magnetics Added To Simscape Pneumatics Added To Simscape Simscape Language Introduced SimElectronics Introduced

Thermal effects optional ports

Code Generation Advances Simscape Diagnostics Improvements

Simscape Introduced SimHydraulics Introduced SimDriveline Introduced SimMechanics Introduced SimPowerSystems Introduced

1998

2000

Simscape-Based Library Introduced

SolidWorks Translator

ProEngineer Translator

Electric Drives Library Introduced

2002

2004

2006

3-D Visualization Improvements AutodeskTranslator Ideal Switching Algorithm Introduced

2008

2010

Second Generation Technology Intf. Elements Editing Modes

2012

2014 4

Optimize System-Level Performance

s3

Controller 

System

Sensors

s2

Actuators

+ u

s1

y

Plant

Simulating plant and controller in one environment allows you to optimize system-level performance. – Automate tuning process using optimization algorithms – Accelerate process using parallel computing

5

Detect Integration Issues Earlier

s2 s3

Controller 

Plant Model System

Sensors

+ u

s1

Actuators

Plant Specification

y

Plant

Controls engineers and domain specialists can work together to detect integration issues in simulation –

– –

Convert plant models to C code for hardware-in-the-loop tests Distribute models to other internal users without extra licenses Distribute models to external users while protecting IP 6

Modeling & Simulation Adoption

Model-Based-Design Adoption Scenarios

Requirementsbased V&V (“connecting” models to requirements, testing reqs in sim, and modeling the requirements)

System Simulation

Virtual Verification & Validation (using Simulation or Analysis with system models & requirements)

Closed-loop Simulation

(either simple plant models or full system models)

Graphical Specs

System Validation (re-application of tests or test plan from simulationbased testing to hardware-based testing, easy comparison of results from Early Testing to Hardware Testing)

Hardware-in-theLoop

Simulation-based Development

(plant-model code gen, maybe prototype codegen too)

(representing and analyzing the system behavior, then generating code for the portion of interest)

Design Prototyping

Graphical Programming

(algorithm code-gen for HW-based prototyping)

Algorithm Modeling (no plant models)

Simulation

Fully-leveraged Model-Based Design

Real-Time Testing

(graphically representing the algorithm and generating code)

Production

Code Generation Adoption 7

Relative cost to fix an error

What is the Most Expensive Project Stage to Find Errors In? Errors introduced early but found late in the process are expensive to fix!

Errors introduced in: coding phase design phase requirements phase

Requirements phase

Design phase

Coding phase

Testing phase

Project phase where error is fixed Source: Return on Investment for Independent Verification & Validation, NASA, 2004. 8

Start Testing on Day One RESEARCH

REQUIREMENTS

DESIGN Environment Models Thermal

Electrical

Supervisory Logic Control Algorithms

IMPLEMENTATION C, C++ MCU

DSP

VHDL, Verilog FPGA

ASIC

TEST & VERIFICATION

Mechanical

Structured Text PLC

TEST SYSTEM

INTEGRATION 9

Early Verification of Concept • Predict dynamic system behavior by simulating - Less physical prototypes DESIGN Environmental Models Mechanical

Thermal

Control Algorithms

• Use of simulation results for system design - What / if studies - Short iteration cycles

Electrical

Idea

Supervisory Logic

Simple model

Detailed model

10

Appropriate Methods of Modeling

DESIGN Environmental Models Mechanical

Thermal

Control Algorithms

Electrical

Algorit hm Devel opme Control & filter nt (Simulink) Data Modeli ng

algorithms

Control & Supervisory Logics (Stateflow)

Supervisory Logic

Embedded Digital Software Electronics VHDL, C, C++ Verilog

Electronical, thermal, mechanical systems (Physical Modeling)

MCUDSP FPGA ASIC Integr Reuse ation

of legacy code & engineering data from - Cosimulation Implement V&V - Exiting algorithms in C, MATLAB 11

Integrated Control Design • DESIGN Environmental Models Mechanical

Thermal

Control Algorithms

Electrical

Reuse of the model for extracting plant description directly from model Algorit • Automated creation of a linearized small signal hm Devel equivalent model at selected operating points opme • Interactive and automatic control design nt Data according Linear Control Theory Modeli ng

Supervisory Logic

Embedded Digital Software Electronics VHDL, C, C++ Verilog MCUDSP FPGA ASIC

•

Integr Robust control design by considering converter ation behavior at several operating points in parallel Implement V&V

12

Test and validate in real-time RESEARCH

REQUIREMENTS

DESIGN Environmental Models Mechanical

Electrical

Control Algorithms

Rapid Prototyping of Control Algorithms

•

Fast implementation of algorithms in C & HDL for functional testing in RT

Supervisory Logic

IMPLEMENTATION C, C++

MCU

DSP

VHDL, Verilog FPGA

ASIC

Structured Text PLC

Hardware-In-The-Loop Testing of Plant •

Capability of testing critical scenarios without risk of damaging HW

13

Automatic Production Code Generation RESEARCH

REQUIREMENTS

DESIGN

General • Code generation in C/C++, HDL, IEC61131- Structured Text • •

Control Algorithms

•

Supervisory Logic

• IMPLEMENTATION C, C++

MCU

DSP

VHDL, Verilog FPGA

ASIC

Structured Text PLC

Fast implementation by automatic code generation from models Support of fixed point data format in simulation and code generation Prevention of implementation errors Algorithm development independent of implementation HW

C-Code • Integration of Legacy C/C++-Code • Automated integration with variety of Embedded IDEs and µP/DSP

14

Traceability from Requirements to Code RESEARCH

REQUIREMENTS

DESIGN

•

Environmental Models Mechanical

Linking Requirements with Model Blocks and generated Code Find corresponding locations easily in model and code

Electrical

Control Algorithms Supervisory Logic

IMPLEMENTATION

15

Benefits of Model-Based Design RESEARCH

REQUIREMENTS



Predict system behavior in early development state



Handle system complexity



Short iteration cycles



Less physical prototypes



Fast implementation by automatic code generation



Reuse of test cases

DESIGN Environment Models Thermal

Electrical

Supervisory Logic Control Algorithms

IMPLEMENTATION C, C++ MCU

DSP

VHDL, Verilog FPGA

ASIC

INTEGRATION

Structured Text

TEST & VERIFICATION

Mechanical

PLC

16

Modeling Physical Systems With MathWorks Products Modeling Approaches First Principles Modeling

Programming

Data-Driven Modeling

Physical Networks

Statistical Methods

(Simscape and other Physical Modeling products)

(Model Based Calibration Toolbox)

(MATLAB, C)

Block Diagram (Simulink)

Modeling Language

(Symbolic Math Toolbox)

(System Identification Toolbox)

Neural Networks

(Simscape language)

Symbolic Methods

System Identification

Parameter Tuning

(Neural Network Toolbox)

(Simulink Design Optimization) 17

Thinking outside the block Physical network approach vs. Simulink block diagrams



Simulink Blocks are casual – Transfer functions – Input and output ports (signal flow) – Graphically model system equations



Physical “Blocks” are acausal – Bi-directional energy flow – Domain-specific physical ports (electrical, hydraulic…) – Graphically model system topology

18

Conservation is Physical 

Simscape generalizes conservation laws across physical domains – i = 0 at every junction (where i is the through variable)



Simscape pre-defined domains: Port Type

Across Variable

Through Variable

Electrical

Voltage

Current

Hydraulic

Pressure

Flow rate

Mechanical (rotational)

Angular velocity

Torque

Mechanical (translational)

Translational velocity

Force

Pneumatic

Pressure and temperature

Mass flow rate and heat flow

Thermal

Temperature

Heat flow 19

Simscape

20

SimPowersystems

21

SimElectronics

SimDriveline

SimHydraulics

Enables physical modeling (acausal) of electrical power systems and electric drives

SimMechanics



SimPowerSystems

Introduction to SimPowerSystems

Simscape MATLAB, Simulink

25kV





Electrical system topology represented by schematic circuit Used by electrical, system and control engineers to develop plant models and test control systems

Breaker

2250 HP Load

3.125 MVA, 2.4kV

22

SimPowerSystems Key Features      

Comprehensive block libraries for building power system models Detailed models of common AC and DC electric drives Different simulation modes to speed model execution Ideal switching algorithm, enabling fast simulation of power electronics PowerGUI provides convenient tools for common analysis tasks Extensive set of demonstration circuits and systems 23

Working with SimPowerSystems SimPowerSystems is a tool for modeling the generation, transmission, distribution, and consumption of electrical power 

With SimPowerSystems you can: – Quickly build electrical power system models – Model synchronous and asynchronous electric drives – Perform common electrical system analysis tasks – Develop and test controls – Generate code for improved performance 24

Quickly Build Electrical Systems 

Build models that look like an electrical schematic: – Three-phase components – Detailed electric drive models – Flexible AC Transmission Systems (FACTS)

 



Parameterize model using MATLAB® variables Connect to Simulink with sources and sensors Save subsystems for reuse in other models or libraries 25

Model Electric Drives 

Combine power electronics, machine, and control algorithm – GUI to assign key parameters – Common strategies for speed and torque control – Adjustable level of fidelity (detailed, averaged)



Common machine types can be used as motors or generators: – – – –

Permanent magnet Synchronous, asynchronous Induction Single phase or 3-phase 26

Calculate Model Parameter Values Asynchronous or PMSM Models 

Calculates parameters for equivalent circuit directly from data sheet values

Data Line to Line RMS Voltage Nominal Frequency Full Load Current Full Load Torque Synchronous Speed

Values 400 V 50 Hz 194 A 352 N.m 2982 rpm

– Plot torque speed relationship – Automatically update machine model Compute Block Parameters

Apply to Selected Block

27

Electrical System Analysis 

Quick access to tools for common analysis tasks: – – – – – –



Display steady-state V and I Display and modify initial states Perform load flows Display impedance measurements FFT analysis Report generation

Multiple simulation modes – Discrete and phasor modes enable you to speed up simulations

28

Improve Simulation Performance 

Ideal switching algorithm – Efficiently recalculate state-space matrix based on states of switches – Displays equations to command window



Enables simulation of ideal switches – Faster simulations (explicit solver, larger time steps) – Does not require difficult parameter values (snubber, etc.) – Remove non-ideal effects, making simulation results easier to interpret

A1x + B1u

A2x + B2u

29

Connecting to Simscape 

Electrical connection via interface blocks – Add custom components using Simscape language – Include other domains



Mechanical ports – Synchronous, asynchronous, DC, and PMS machines

30

SimPowerSystems Supports Simscape Editing Mode 

Share SimPowerSystems models with other Simscape users

Model Developer Purchases Simscape and add-on products

– Simulate, analyze, generate code without purchasing extra licenses

Function

Full Mode

Restricted Mode

Add or delete regular Simulink blocks

Yes

Yes

Change Simulink solver, simulate

Yes

Yes

Change numerical parameters

Yes

Yes

Access PowerGUI functions, settings

Yes

Yes

Generate code

Yes

Yes

Add/delete blocks from add-on products

Yes

No

Make or break physical connections

Yes

No

Change block parameterization options

Yes

No

Change Simscape Local Solver

Yes

No

Model using Simscape and add-on products

Model Users Purchases Simscape Add-on product installed, No add-on purchases required

31

Developing Control Systems 



Implement high-fidelity, nonlinear plant models

u

Extract linear model for use with linear control theory

+

s1

s2 s3

Controller

Ax+Bu=0 Root Locus





Explore interaction between control system and plant

Plant

Real Axis

Bode Plot

Frequency

Optimize system performance

32

SimPowerSystems Demonstration PM Synchronous Motor Drive (ac6_example.mdl)

Discrete solver provides fast, accurate results. 33

– Models look like power network schematic – Three-phase, electric drives, FACTS, etc. 

SimElectronics

SimDriveline

SimHydraulics

Enables physical modeling of electrical power systems

SimMechanics



SimPowerSystems

SimPowerSystems Summary

Simscape MATLAB, Simulink

Solvers optimized for fast simulation of high-speed switching electronics – Continuous, discrete, and phasor methods – Ideal switching algorithm



Many analyses are automated – Load flows, FFT analysis, and more



Combining with other physical modeling tools to model multidomain systems 34

ABB Accelerates Application Control Software Development for Power Electronic Controller Challenge Adopt a more efficient development process using tools that accelerate the design of new application software for a high-powered electronic controller for power converters

AC 800PEC controller.

Solution Use MathWorks tools to design and validate their control algorithms while streamlining the application software development process for the controller

Results  Development times and costs reduced  Development process improved  Highly accurate code generated

“Our system engineers can program, simulate, and verify the AC 800PEC controller’s regulation software very rapidly

in MATLAB and Simulink.” Fritz Wittwer ABB Link to user story

35

Alstom Generates Production Code for Safety-Critical Power Converter Control Systems Challenge Design and implement real-time power conversion and control systems for trams, metros, and railways Pendolino tilting train.

Solution Use MathWorks tools for Model-Based Design to design, simulate, and automatically generate production code for safety-critical transportation systems

Results  Development time cut by 50%  Defect-free, safety-critical code generated and certified  Common language established

“When Alstom delivered a Pendolino train to Czech Railways, the railway application was the first with automatically generated code to receive TUV certification.” Han Geerligs Alstom

Link to user story

36

Hydro-Québec Models Wind Power Plant Performance

Challenge Plan the integration of new wind farms into the power system, predict power output, and ensure safe, reliable operation

Turbines on a wind farm.

Solution

“Accurate modeling is essential not

Use MathWorks products to simulate individual wind turbines and wind farms and to generate C code for multiprocessor simulation of entire power systems

only for planning investments but

Results

tools, we can simulate power

 Simulation speed increased to real time  Equipment needs accurately predicted  Dynamic simulations enabled

also to detect situations that can cause an outage. With MathWorks electronics, mechanics, and control systems in one environment, and

our models respond like the turbines we have in the field.” Richard Gagnon Hydro-Québec

Link to user story

37

What is xPC Target?

© 2012 The MathWorks, Inc. 38

Model-Based Design: Early V&V is the Key REQUIREMENTS

Verification & Validation Design Environment

Test

Integration Design Testing Verification

Testing (HIL)

Controller Code Verification

Implementation Embedded Software

C, C++ MCU

DSP

39

Real-Time Testing Scenarios: Functional Rapid Prototyping

Code Generation

Execution Host/Target Real-time

Wiring and Signal Conditioning

Real-Time Target Computer

Production Plant Hardware

40

Real-Time Testing Scenarios: Hardware-in-the-Loop (HIL) Simulation

Code Generation

Execution Host/Target/Target Real-time

Code Generation

Wiring and Signal Conditioning

ECU or MicroController

Real-Time Target Computer

41

What is xPC Target?

1

2

Host computer with MATLAB

xPC Target on Target Computer

3

Ethernet or RS 232

 Environment allows the real-time execution of Simulink models on a separate PC-based target computer 42

What is xPC Target? xPC Target on Target Computer

Host computer with MATLAB Host computer with MATLAB

Ethernet or RS 232xPC Target

on Target Computer

Ethernet or RS 232

 Environment provides interactive access between the real-time application and the host computer  Allows live parameter tuning, control from the original Simulink model and offline analysis support in MATLAB.

43

What is xPC Target?

2

Host computer with MATLAB

xPC Target on Target Computer Ethernet or RS 232

4

2 3 1

 Environment provides interactive access between the real-time application and the host computer  Allows live parameter tuning, control from the original Simulink model and offline analysis support in MATLAB.

44

What is xPC Target? Host computer with MATLAB

xPC Target on Target Computer Ethernet or RS 232

 Environment provides numerous I/O device driver blocks  Blocks are easily configurable within the Simulink model and communicate with actual hardware in real-time. 45

What is xPC Target? xPC Target on Target Computer

 Environment provides numerous I/O device driver blocks  Blocks are easily configurable within the Simulink model and communicate with actual hardware in real-time. 46