Developing DC-DC Converter Control with Simulink

Vasco Lenzi
Senior Application Engineer

© 2020 The MathWorks, Inc.

1
Key Takeaways

▪ Graphical programming across our solutions is intuitive and powerful

▪ State-of-the-art technologies facilitate the design and verification of

complex systems developed by multidisciplinary teams

▪ Find design errors early and cut down development cost while increasing
delivered quality.

2
Our Project Today
DC/DC LED Developer's Kit

LED Head Lamp

3
ZKW Lichtsysteme GmbH

Rapid Control Prototyping with Simulink Real-Time and Speedgoat:

▪ Design control algorithms for an innovative LED headlamp projection technology changing its
illumination dynamically
▪ Seamless integration into MathWorks Tools
▪ Faster time to market

"Model-based design itself has proven to be very flexible, powerful and

efficient for our purposes. Using the Mobile real-time target machine
from Speedgoat, we were able to completely redesign a functional
prototype based on a model and verify it during an afternoon session."

- Matthaeus Artmann, Manager Electronics Engineering Pre- and

Module Development, ZKW Lichtsysteme GmbH

Official Speedgoat User Story

https://www.youtube.com/watch?v=wAk9e5w0dSg

4
Developing DC-DC Converter Control with Simulink

▪ Model the converter and calculate the most efficient operating region

▪ Determine power losses and the thermal behaviour of the converter

▪ Design control algorithm based on time/frequency domain specification

▪ Design supervisory logic and implement unit testing

▪ Implement power electronic controls on an embedded platform

5
6
Developing DC-DC Converter Control with Simulink

▪ Model the converter and calculate the most efficient operating region

▪ Determine power losses and the thermal behaviour of the converter

▪ Design control algorithm based on time/frequency domain specification

▪ Design supervisory logic and implement unit testing

▪ Implement power electronic controls on an embedded platform

7
 Modeling the converter

SEPIC

Load (LEDs)

Power Supply
8
Modeling Approaches

First Principles Data-Driven

Physical Networks Statistical Methods
Programming System
Identification
Block Diagram

Modeling Language Neural Networks

Symbolic Methods Parameter Tuning Measured

Model

9
Simscape Products

▪ MATLAB and Simulink provide

foundation for technical computing
and algorithm development

▪ Simscape platform
– Simulation engine and custom diagnostics
– Foundation libraries in many domains
– Language for defining custom blocks

▪ Simscape add-on libraries

10
What’s new in Simscape Electrical

▪ Parametrized Stepper motor block

▪ Parametrized Battery block

▪ Faults
– Dynamic Load from DC or AC supply
– Constant Power Load
– Delta-Connected Load, Wye-Connected Load

11
Developing DC-DC Converter Control with Simulink

▪ Model the converter and calculate the most efficient operating region

▪ Determine power losses and the thermal behaviour of the converter

▪ Design control algorithm based on time/frequency domain specification

▪ Design supervisory logic and implement unit testing

▪ Implement power electronic controls on an embedded platform

15
 ee_switching_power_supply_thermal
Recap: What have we seen?

• Create heat maps

• Reuse it in extremely fast thermal-focused
model for cooling sizing and control

ee_getpowerlossSummary

16
 Convert SPICE models into Simscape components
Netlist

subcircuit2ssc

17
Simscape Electrical
SPICE Models

▪ Manufacturer-specific MOSFETs

▪ Additional transistor capacitance models

in SPICE NMOS and SPICE PMOS blocks
– Meyer gate or charge conservation

▪ Conversion Assistant supports

table SPICE function

▪ Validate MOSFET conversions by

generating characteristics and
comparing with LTspice
18
Developing DC-DC Converter Control with Simulink

▪ Model the converter and calculate the most efficient operating region

▪ Determine power losses and the thermal behaviour of the converter

▪ Design control algorithm based on time/frequency domain specification

▪ Design supervisory logic and implement unit testing

▪ Implement power electronic controls on an embedded platform

19
Voltage Controller Design

▪ Requirement
– Implement voltage controller and tune it

▪ Approach
– Create transfer function equivalent model
– Tune controllers based on requirements

20
What have we seen?

• Tuning based on System Identification

• Works with any topology

21
Average switch option for converters and choppers

▪ Faster simulation by using modulation

signal or undersampling as gate signal
– Bidirectional DC-DC Converter
– Boost Converter
– Buck Converter
– Buck-Boost Converter
– Converter (Three-Phase)
– Four-Quadrant Chopper
– One-Quadrant Chopper
– Three-Level Converter (Three-Phase)
– Two-Quadrant Chopper

22
Developing DC-DC Converter Control with Simulink

▪ Model the converter and calculate the most efficient operating region

▪ Determine power losses and the thermal behaviour of the converter

▪ Design control algorithm based on time/frequency domain specification

▪ Design supervisory logic and implement unit testing

▪ Implement power electronic controls on an embedded platform

23
Recap: What have we seen?
Stateflow

24
Recap: What have we seen?
Simulink Test

✓ Create test harnesses

and test cases

✓ Group into suites and

test files

✓ Execute individual or
batch

✓ View result summary

✓ Analyze results

✓ Archive, export, report

25
Developing DC-DC Converter Control with Simulink

▪ Model the converter and calculate the most efficient operating region

▪ Determine power losses and the thermal behaviour of the converter

▪ Design control algorithm based on time/frequency domain specification

▪ Design supervisory logic and implement unit testing

▪ Implement power electronic controls on an embedded platform

26
Automatic Code Generation

▪ Requirement
– Generate target-aware, efficient C-code

▪ Approach
– Model elaboration for C-code generation
– Create a first configuration set with Embedded Coder Quick Start
– Build the code, automatically generate reports

27
 Deploy to Any Processor with Best-in-Class Performance
Models in MATLAB and Simulink can be deployed on embedded devices, edge
devices, enterprise systems, the cloud, or the desktop.

Deployment CPU

Embedded devices
Automatic
Code
Generation
GPU
Enterprise systems

Edge, cloud,
desktop FPGA

PLCs
28 28
 MATLAB Connects to Your Hardware Devices

Instrument Control
Oscilloscopes, Signal Generators,
Lab Instruments

Data Acquisition
Plug-in data acquisition devices, I/O
boards and sound cards

Image and Video Acquisition

Industrial and scientific cameras

Digital Networks
OPC, CAN, J1939, and XCP protocol
devices

Hardware support packages

Built-in and downloadable support for a wide range of devices
and development boards
29
Project and File Management Included in
Simulink since
2017b!

Simulink Project Upgrade Simulink Graphical Model Comparison&Merge

30
In-the-Loop Verification Methodologies
Software-in-the-Loop

Is the generated
Compiled
Algorithm Environment code functionally
SIL equivalent to the
model ?

31
In-the-Loop Verification Methodologies
Processor or FPGA-in-the-Loop

Is the generated
Algorithm Environment code functionally
equivalent to the
model ?

Non-Real-Time functional
verification of the algorithm
component, C or HDL

32
 In-the-loop verification methodologies
Hardware-in-the-Loop: “HIL”

Does algorithm
perform well on
Algorithm Plant Model actual device
with true
latencies?
Code Code

Production embedded target:

Real-Time Machine
Structured Text, VHDL, C/C++
Harness eg “Speedgoat”

33
 About Speedgoat

▪ A MathWorks associate company, incorporated in 2006

by former MathWorks employees. Headquarters in
Switzerland, with subsidiaries in the USA and Germany

▪ Provider of real-time target computers, expressly

designed for use with Simulink

▪ Real-time core team of around 200 people within

MathWorks and Speedgoat. Closely working with the
entire MathWorks organization employing around 5,000
people worldwide

34
Hardware-in-the-Loop Simulation of SEPIC Converter
Linearized SEPIC used to
deploy to FPGA I/O module
Target Computer via Simcape HDL Workflow
Development with FPGA I/O Module
Computer
Model sample time = 500 nanoseconds (2 MHz)

DAC - Analog output Device Under Test

Xilinx Kintex®-7 FPGA • ~ 500 ns settling time
• 16-bit
Output voltage
Floating-point • ±10V voltage range
support

PWM Capture - 200 MHz PWM signal

• 5 nanosecond resolution

100 kHz
35
Hardware-in-the-loop simulation of SEPIC converter

36
37
With Simulink and Model-Based Design
Use Simulation Models to Generate Production Ready Code

“With Model-Based Design, our developer

productivity is easily increased tenfold. Simulation
and code generation enable us to turn changes
around quickly and eliminate human errors in
coding. Our algorithms typically work the first time,
so we no longer waste a big part of our development
cycle debugging code.”
A cabinet of Power Electronic
Building Blocks (PEBBs). Dr. Robert Turner, ABB link

38
Key Takeaways

▪ Graphical programming across our solutions is intuitive and powerful

▪ State-of-the-art technologies facilitate the design and verification of complex

systems developed by multidisciplinary teams

▪ Find design errors early and cut down development cost while increasing
delivered quality.

39
 Visit the Power Electronics Control Community on
MATLAB Central to find Models, Answers, and How-to Videos

https://www.mathworks.com/matlabcentral/topics/power-electronics-control.html

40
Q&A and Conclusion

▪ Questions: [email protected]

41