|
Course
Highlights
This course is designed to cover the basic theory of control design to hardware implementation using the Xilinx FPGA. The course is divided to three modules which are the theory, software modeling and hardware implementation.
The theory covers the fundamentals of control system to system modeling. The MATLAB® and Simulink® software will be used for system analyses to confirm the studied theories and to simulate a controlled system dynamically. Finally, the steps to implement the ideally simulated controller model in hardware will be shown; the challenges in transferring an ideal model to a realizable hardware will be discussed. The Xilinx FPGA, Spartan 3a evaluation board, will be used for the controller implementation as it provides various peripherals and extended modules to interface with other hardware.
Course
Objectives
- To provide participants with a comprehensive approach towards control system design and analyses using MATLAB- and Simulink- Control System libraries.
- To present a systematic approach of designing a controller suitable for hardware implementation by taking participants from theory to practice.
Who
Must Attend
Engineer, researchers, scientists, and managers who are involved in control engineering design and problem solving. It is also strongly recommended for those who would like to establish and strengthen their foundation in Control Engineering.
Course Benefits
Upon the completion of the course, the participants will gain a comprehensive understanding on utilizing the Control System Toolbox, Stateflow, Real-Time Workshop and FPGA to design and develop control system using a variety of design methodologies.
Prerequisites
Attended "Comprehensive MATLAB"
and "Comprehensive SIMULINK" or equivalent experience using MATLAB and
SIMULINK, and an understanding of terminology and concepts
related to common control systems.
Course
Outline
The sections of the course are outlined below:
The sections of the course are outlined below:
1. Modelling with MATLAB and Simulink
This section reviews the main features of MATLAB and SIMULINK to be used throughout the course. Basic data manipulation and data visualization capabilities of MATLAB will be reviewed. The capability of building models and creating subsystems using SIMULINK will be revisited. Common functions and arithmetic expressions relevant to the Control System Design will be highlighted. Furthermore, the structure and cell data types are presented as a precursor to objects, which form a central part of the Control System Toolbox. The model for the magnetic levitation device (the process to be controlled) is developed.
2. Poles, Zeros and Time Response
Poles and zeros play a very important role in the dynamic response of any control system. In this section, the significance of poles and zeros in relation to the dynamic response of a control system will be presented. Control System Toolbox functions that can transform a system model from either a transfer function or a state-space model to a pole-zero form will be discussed. MATLAB functions to compute and plot the dynamic response of a control system represented by a pole-zero form will also be highlighted. Furthermore, techniques to obtain dynamic responses for linear systems as well as nonlinear systems will be presented.
3. Overview of the Control System Toolbox
Building on the foundation of MATLAB, the Control System Toolbox provides a collection of algorithms that simplify common control system design, analysis and modeling tasks. This section presents the main features of the toolbox and a good understanding of how to use them effectively. Representations of continuous-time and discrete-time systems in transfer functions, state-space model and poles-zeros form, functions to change concept of poles and zeros, inter-relationships of the three forms, functions for time response, frequency responses, root locus plots, pole placement, optimal control and estimation will be presented.
4. Controller Design Using MATLAB and SIMULINK
Building on the first three sections of the course, in this section, a control law for the magnetic levitation device is developed. Important design considerations such as signal filtering and the dynamics of the A/D
and D/A channels are included. Several techniques for model liberalization are shown and a simple PID controller derived. Using the Nonlinear Control Design Toolbox, the parameters of the PID gains are tuned.
5. Code Generation with Real-time Workshop
Real-time Workshop is the tool for automatically converting a SIMULINK model into equivalent C code. This fully customizable process reduces the need for the time-consuming (and often error prone) task of handling coding control laws. Code will be generated for several purposes, including fast batch mode simulation runs (for parameter tuning), for encrypting all or part of a model and for real-time implementation. Participants will have the opportunity to see how to customize the code generation process and generate code for the controller they have designed.
6. Applying Finite State Machine Modeling with Stateflow
Understand how to use Stateflow to model finite-state machine theory and supervisory logic. The course discusses how to interact with Simulink, and graphically build flow diagrams and functions. Code generation and sending data out of Stateflow are briefly mentioned in this course as well.
7. Real-time Implementation using RTW
PC-based processors are becoming increasingly popular for data acquisition and real-time proto-typing of control systems. In this section, the essential features of Real-Time Windows Target, one of the MathWorks solutions for PC-based real-time implementation, is presented.
8. Real-time Implementation using FPGA
FPGA has been emerged as a viable semiconductor to ASIC. The Xilinx cheapest FPGA Spartan 3a will be used in the hands-on implementation of the controller in real-time – providing higher processing rate as compared to the PC based processor. However, the system development in FPGA requires knowledge in low-level technical design such as VHDL or Verilog programming. Nevertheless, the features of model-based design using Simulink automated generated VHDL code will be shown, which eases the design challenges and speed up design process.
|
