With the growth of microcomputer capabilities, control engineering has witnessed a significant shift towards digital implementation of digital controllers. Many modern industrial and commercial control systems employ digital computers. With the use of digital computers in the loop, one can readily generate C code from graphical simulation tools such as Simulink block diagrams for real-time controller implementation. This has provided the impetus for establishment of a real-time instructional control laboratory at Milwaukee School of Engineering. In this paper the development of this innovative integrated real-time control system laboratory will be described. Next modeling, simulation, controller design and implementation of a few typical laboratory experiments and projects of different complexity are presented. This paper also reports on Simulink modeling of the nonlinear inverted pendulum with some research results for different swing-up controllers. Introduction With the great advances in microelectronics and high-performance data acquisition and control boards, the control of modern industrial and commercial systems with a digital computer is becoming more common. To bridge the gap between academics and industry it is essential to supplement the teaching of control system courses by developing and incorporating into the curriculum a realtime control system laboratory that will allow students to design, implement, and test their own control systems. For a control education laboratory to work effectively and efficiently it should be based on well established systems that provide a variety of experiments that span the spectrum of topics from classical control to optimal, digital, robust, and nonlinear control. If such a system is in place then it becomes feasible for the research professor to concentrate on ”build your own” experiments. Or, more appropriately, it gives graduate students an opportunity to build their own open-ended exploratory experiments. The Electrical Engineering and Computer Science Department at the Milwaukee School of Engineering has recently established a modern real-time control system laboratory. This laboratory was developed to enhance the quality of undergraduate education in control systems and to reinforce the concepts that are covered in the lectures. This real-time control laboratory also opens up research activities for graduate student projects. The undergraduate experiments give students the opportunity to tie theoretical classroom concepts with practical hands-on experiments. The objective is to provide undergraduate students with computer-based control and introduce them to concepts such as client-server environment, visual instrumentation, automatic code generation, and the modern techniques needed for the design and implementation of automatic control systems. In addition, the use of the Matlab/Simulink combination for analysis and design of modern control systems and the implementation of physical control systems has allowed coverage of topics in undergraduate curriculum that traditionally are considered advanced topics. Proceedings of the 2004 American Society for Engineering Education Annual Conference & Exposition Copyright 2004, American Society for Engineering Education P ge 938.1 The remainder of this paper describes the equipment and software installed in the laboratory, provides an overview of the experiments in the introductory junior-level controls course with details on the rate feedback experiment, and provides an overview of the projects in the senior-level controls course with details on the inverted pendulum project. Both courses are taught on an 11 week quarter system. The laboratory has 10 fully equipped stations with two students working at a station as a team for a capacity of 20 students per section. In the required introductory junior-level course, a two hour per week laboratory session is conducted in addition to three 50 minute lecture sessions per week. In the elective senior-level course, there are three 50 minute lecture sessions per week and no scheduled laboratory sessions. Students work on two different projects throughout the quarter. At MSOE, there are no graduate teaching assistants and the course instructor delivers lectures and directly supervises laboratory sessions. This provides a tight coupling between the lectures and the laboratories. The controls courses as taught at MSOE would require significant modification if the course instructor did not directly supervise the laboratory sessions and if section sizes were much greater than 20 students. More detailed information about these projects and the laboratory manual used in the first course can be found at http://www.msoe.edu/ ̃saadat/control_system_instructional_lab.htm Equipment and Software The laboratory is equipped with state-of-the-art computer aided control equipment for analog and digital control systems simulation, analysis, design, and implementation. A conceptual layout of the equipment is shown in Figure 1. The analysis and design of control systems is carried out via MATLAB and Control System Toolbox, which provide students with immediate access to an extensive range of analysis and design tools. The graphical simulation tool Simulink is used for modeling, design and simulation together with Real-Time Workshop to generate C code from Simulink models for control of servo systems in real time. Finally, Quanser’s WinCon software invokes a C compiler to create the executable control code; downloads the executable code to a client controller computer; provides data transfer between Simulink and the servo system; enables on-the-fly controller tuning; and plots data in real-time. The WinCon software consists of two programs: WinCon client and WinCon server. The WinCon client is installed on the control host computer, and WinCon server is installed on the student’s laptop. The student’s laptop also contains MATLAB, Simulink and the Real-Time Workshop software. Figure 1. Real-Time Control System Laboratory Ten Allen Bradley industrial micro PC’s provide the WinCon client controllers. Each PC contains the Quanser ISA MultiQ-2E Data Acquisition and Control Board enabling real-time processing and storage of multi-channel analog input and output data. Each station is equipped with a Quanser UPM1503-24V Power Module and SRV02-ET Servomotor that includes encoder, tachometer and potentiometer. Several modular mechatronics components provide the capability for design and Proceedings of the 2004 American Society for Engineering Education Annual Conference & Exposition Copyright 2004, American Society for Engineering Education P ge 938.2 control of systems with a wide variety of plant dynamics for several experiments of different degrees of difficulty. Also at each station there is an Agilent 54600 oscilloscope, HP 33120A Function Generator, HP E3631A Power Supply, and an HP 34401A multimeter, all connected to a General Purpose Interface Bus (GPIB). The students perform all their design work on their laptops. They connect their laptops to the host computer via an Ethernet cable. WinCon server enables students to connect to the desired workstation over the internet and generate the real-time code on their laptops. The real-time code is downloaded automatically to the host computer. The control is executed on the host computer and data is streamed to the laptop for on-the fly controller tuning and real-time plotting. Laboratory Experiments An innovative approach is to integrate the real-time control laboratory in the first control system core course where some laboratory sessions can be used for instruction as well as hands-on experience of lab work. The experiments are designed in a manner which allows a reinforcement of the concepts that are covered in the lecture. The developed modular experiments and laboratory manual allow the student to become familiar with the commonly used devices such as sensors, transducers, signal conditioning circuits, microprocessor interface, controller design, implementation and test on the real-time hardware. In the first session students learn the basics of the software and hardware used for implementation of control systems. The second session is devoted to operational amplifiers; A/D-D/A converters; and emulation and implementation of typical compensators. The next two sessions focus on modeling, an orderly approach to digital simulations, and investigation of the effect of damping on the system response. Case studies include the third-order Butterworth filter, a mechanical system, a moving coil actuator, a simple pendulum, a nonlinear system and the inverted pendulum. In the six remaining sessions students perform the following experiments: • Position Control I (Rate Feedback) • Velocity Control I (Velocity Feedback) • Position Control II (PD Controller) • Position Control III (Phase-lead Controller) • Ball and Beam Project For each experiment students develop the system model, design the controller based on reasonable performance criteria, and simulate their control system in Simulink. Next, students construct an implementation diagram, build and download it to one of the client workstations, and establish if the control system meets the design specifications. One of the above experiments, Position Control I (Rate Feedback), is briefly presented here. Position Control I (Rate Feedback) Position control systems are used extensively in industrial applications such as robotics and drive control. Modern position control systems are achieved with precision using incremental encoder sensors. In this experiment students design and implement a position control system for low frequency square wave input. The objectives of this project are: • To obtain the servo plant model • To design a position control system such that the output angle tracks a commanded position using position and velocity feedback and determine the feedback gains to achieve the given time-domai
[1]
Richard T. O'Brien,et al.
A Novel Approach to a Control Systems Laboratory
,
2003
.
[2]
Bin Yao,et al.
The Educational Impact of a Gantry Crane Project in an Undergraduate Controls Class
,
2002
.
[3]
Hartono Sumali,et al.
Teaching Instrumentation and Data Acquisition at the Federal University of Viçosa in Brazil
,
2002
.
[4]
Z. Luo.
Direct strain feedback control of flexible robot arms: new theoretical and experimental results
,
1993,
IEEE Trans. Autom. Control..
[5]
Charles Knight.
Modern Engineering Laboratories That Deliver
,
2002
.
[6]
J. A. Piepmeier,et al.
Developing Feedback Control Prototypes Using A Real Time Simulink Environment
,
2003
.
[7]
K.S. Rattan,et al.
Identification and control of a single-link flexible manipulator
,
1988,
Proceedings of the 27th IEEE Conference on Decision and Control.
[8]
Jamal Daafouz,et al.
Robust control of a flexible robot arm using the quadratic d-stability approach
,
1998,
IEEE Trans. Control. Syst. Technol..
[9]
D. C. Kennedy,et al.
Integrating the Internet, multimedia components, and hands-on experimentation into problem-based control education
,
2000,
30th Annual Frontiers in Education Conference. Building on A Century of Progress in Engineering Education. Conference Proceedings (IEEE Cat. No.00CH37135).
[10]
John Spinelli.
The Signals And Systems Toolbox: Comparing Theory, Simulation, And Implementation Using Matlab And Programmable Instruments
,
2003
.
[11]
D.S. Bernstein,et al.
Experiments for control research
,
2003,
IEEE Control Systems.