EEE 481 Computer Controlled Systems COURSE DESCRIPTION (EXTENDED) The ever-increasing availability of reliable, low-cost, high-performance computing hardware has been one of the key drivers behind the proliferation of computer-based control of processes. Such controllers can perform a variety of complicated tasks that increase the overall process performance and repeatability, decrease the operating cost, meet stringent safety and environmental constraints, and vastly improve the human user interface. Typical examples of computer controlled systems range from the mundane to the exotic: Modern washers use computer control to provide a cleaner laundry, using less energy and less water. Computers enable car engines to deliver more horsepower, with lower fuel consumption and emissions, and allow sophisticated diagnostic tests. Industrial controllers enable more efficient manufacturing, while state-of-the-art control in aerospace applications is responsible for the extraordinary performance of fighter aircraft, and the cost efficiency of commercial airliners. In addition to the advances in hardware, the software evolution has played a key role in enabling these applications. The complexity of the modern controllers is such that “controller evaluation, commissioning, and maintenance” have taken new meaning. New concepts like “rapid prototyping” have emerged. A modern process controller can deliver the computing power of an old workstation and occupy ¼ of the volume of a notebook PC. Such controllers can now be “embedded” in a process tool and perform a variety of tasks to ensure process reliability and accurate diagnostics. By doing so locally, independent of a central monitoring computer, they free up resources and improve the fault-tolerance of the overall system. In return, they require more elaborate communications and networking. The design and implementation of a modern control system should still address the classic fundamental problems of sensors and measurements, actuators, and feedback. But it must also account for the issues associated with real-time operation, computer hardware and software, discretization and quantization, interfacing and communications. State-of-the-art applications, for which a high degree of integration and optimization of every aspect of the control system is critical, require significant investments and large groups of experts. However, it is only in the recent years that complete control solutions of medium-large-size problems are within the capabilities of a small R&D group or even an individual. There are primarily two reasons for this: One is that the computing power (speed/memory) of general-purpose computer platforms now exceeds the needs of typical control algorithms. The other is the ability of modern CAD software (e.g., MATLAB) to generate stand-alone executables from codes written in a high-level language. The former obviates the need to use highly specialized hardware and avoids the associated development overhead. The latter allows the programming of sophisticated control algorithms in a compact form that can be easily debugged, improved and maintained. The combination of both of these technology advancements has resulted in a quantum breakthrough giving practical meaning to the “rapid prototyping” of sophisticated control algorithms. This “Computer Controlled Systems” course aims to address the variety of issues arising in the implementation of computer-based controlled systems in an integrated and simplified manner. By focusing on the use of PC104 boards and MATLAB xPC Target toolbox, it provides a global perspective of the entire process of implementing and testing embedded controllers, but avoids time-consuming dwellings in less-often encountered details. Smallscale but representative projects form the core of the instruction and provide hands-on expertise. Course Description (20 word, catalog): Implementation of computer-based, embedded, control systems using MATLAB xPC Target toolbox. Small-scale, representative projects demonstrate theoretical issues and provide hands-on expertise. Prerequisite: EEE 203, EEE 221(or equivalent) Textbook: M. Chidambaram, Computer Control of Processes. Alpha Science Intl. Ltd, Pangbourne, 2002. (ISBN 1-84265-063-7 or 0-8493-1010-5) Franklin, Powell, Workman, Digital Control of Dynamic Systems. 3rd ed. Ellis-Kagle Press, Half Moon Bay, Ca. C.L. Phillips and H.T. Nagle, Digital System Control Analysis and Design Prentice Hall, 3rd Ed. Supplemental Materials: Class notes and software distributed by the instructor. (See course web page for additional references.) Coordinator: K. Tsakalis Course Objectives: 1. Students are familiar with the most common elements of computer control: sensors, sampling, control algorithms, actuators. 2. Students understand the basic problems in computer control of processes: principles of feedback and feedforward control, sampling, quantization, real-time operation. 3. Students are familiar with computer software to implement embedded control systems (e.g., MATLAB, Real Time Workshop, xPC Target). Course Outcomes: 1. Students can discuss the principles of operation of common sensors and actuators, A/D-D/A converters. 2. Students can state and apply basic definitions in measurements. 3. Students can apply standard design techniques for common control algorithms (PID, feedforward). 4. Students can discuss and analyze issues related to controller discretization and signal/parameter quantization. 5. Students can use computer software to implement embedded controllers. Course Topics: 1. 2. 3. 4. Examples of computer controlled systems Open-loop (feed-forward) and closed-loop (feedback) control Position control. Velocity control. Temperature control. Flow control. Pressure control. Voltage, current, power control. Light intensity control Instrumentation Measurement basics: Error analysis and statistics. Validity, Repeatability, Accuracy, Precision, Resolution. Static-Dynamic errors Electrical Measurements Voltmeters, Ammeters. Potentiometers, Position/Displacement Transducers, Accelerometers, Tachometers. Bridge circuits. Thermocouples, Thermistors. Strain gauges, Force/Pressure Transducers, Flow meters. Light/Optical sensors Digital Sensors. Counters, Trigger circuits. A/D, D/A converters Real-Time and Discretization Issues Real-time operating systems (introduction), Interrupts. Timing diagrams, Event-sequence descriptions. Computer data acquisition. Discretization errors, Sampling and reconstruction. Anti-aliasing filters, Analog implementations. Quantization errors. Software and Hardware Platforms MATLAB/SIMULINK and other high-level development environments. Embedded Controllers and stand-alone applications; xPC Target Toolbox, Auto-code generation. The PC104 platform. 5. 6. 7. Digital communications principles (computer-sensor, computer-computer). RS232, GPIB, SECS, Ethernet protocols. Actuators Electrical, Pneumatic, Hydraulic actuators. Solenoids, Relays. SCRs, TRIACs. Motors (DC, AC, Stepping). Current-to-pressure actuators. Valves, Mass-Flow Controllers. Control Algorithms and Procedures Event-driven (discrete-state) control. Ladder diagrams, Programmable Logic Controllers Sampled-data (continuous-state) control. Dynamical Systems descriptions, state-space concepts, discretization. Feedforward controllers, dynamic inversion. Feedback Systems: Basic Properties, Stability, Sensitivity. On-Off (2-state) controllers. Chattering, Cycling. Classical simple dynamic feedback controllers. PID control principles. Discrete-time implementations, quick tuning procedures. What’s next? (Overview of advanced topics) Signal Processors. FPGAs. Smart Sensors. Data Mining. Data Fusion. Control system architectures: Distributed, Decentralized, Cascade, 2-DOF Controllers. Modeling and Approximations. System Identification, Neural-Networks. Optimal Control Systems (Feedback, Feed-forward). Adaptation and Learning. Intelligent Systems. Computer Usage: Exercises and demonstrations using MATLAB/SIMULINK and embedded controller hardware. Laboratory Experiments: 1. Familiarization with basic hardware connections and procedures to create real-time executables. 2. RS-232 serial port asynchronous communication. 3. Target-Host communications and MATLAB programming. 4. A/D-D/A converters. 5. Modeling and implementation of a virtual heat transfer experiment; PID control. 6. Implementation of simple least-squares and Kalman filter estimators for system identification. 7. Modeling and implementation of a virtual water level control experiment. 8. Modeling and implementation of a virtual inverted pendulum experiment; more complicated controller design and communication requirements. Course Contribution to Engineering Science and Design: EEE 481 emphasizes engineering design by using open-ended exercises and hardware/software implementation of theoretically derived Emphasis is placed on integrating verious components Since there are many possible solutions to such a problem, students are able to consider design tradeoffs and issues involved in practical implementation. Course Relationship to Program Objectives: A.1 [3] Students are exposed to modern computational techniques for designing feedback and feedforward control systems. They are also exposed to modern CAD tools (e.g., MATLAB, RTW, xPC-Target) that have been used extensively during the last decade in industry and academia for modeling and implementation of control systems. A.2 [2] The operation of basic sensor and actuators, analog-to-digital and digital-to-analog conversions,and common computer communication protocols, are reviewed to enable their usage in control system implementations. A.3 [1] The students use modern CAD tools that reflect industrial trends in embedded systems applications. C.1 [2] Students receive hands-on experience in modern CAD, analysis, and implementation tools that are extremely desirable in the constituent industry. C.2 [1] Hands-on experience is a desirable quality in applying to graduate programs. D.1 [2] Students learn the advantages and limitations of various hardware and software components of control systems. They are using this knowledge to provide integrated control system solutions for virtual properties. D.2 [1] Students obtain models for various processes and create virtual (computer simulated) experiments, for which they then design control systems. D.3 [1] The translation of physical problems into an abstract but rigorous and quantitative mathematical framework enhances the understanding of the physical phenomena. The creation of virtual experiments also requires a deeper understanding of model limitations. D.4 [3] Students use MATLAB extensively for simulation and real-time implementation. D.5 [1] Through A.1, A.2, D.2, D.3 Person preparing this description and date of preparation: Kostas Tsakalis, April 2005.