An Inexpensive Control System Experiment: Modeling, Simulation, and Laboratory Implementation of a PID Controller-Based System

This paper presents a classroom-proven control system experiment that conveys the fundamental concepts of designing a PID controller based closed-loop system. The laboratory experiment presented herein provides an opportunity for students to model, design, simulate, and implement a complete feedback control system in a very inexpensive way by using only a couple of quad op-amp ICs and a few discrete resistors and capacitors. Students are able to design various controller configurations (P, PD, PI, and PID) and investigate the effects of proportional, integral, and derivative gains on system performance including steady state error, damping ratio, overshoot, rise time, time constant, settling time, frequency of oscillation, and system stability. This laboratory experiment emphasizes developing an intuitive understanding of PID controller concepts grounded in theory and design, yet not too mathematical in nature to impede learning for many of the engineering technology students we serve. Introduction Teaching a control systems course can easily become too mathematical for many engineering technology students. This is mitigated primarily by using simulation and hands-on experiments in support of developing and strengthening basic control concepts such as system modeling, steady state and dynamic performance characterization, time domain and frequency domain analysis, specification based controller design, and system stability. A specific topic of significant importance is the conceptual, mathematical, practical, and intuitive understanding of proportion-integral-derivative (PID) controller design and its hardware implementation. This paper presents a very inexpensive and highly effective PID controller laboratory experiment whereby students analyze and mathematically model a mass-spring-damper system, followed by simulation and op-amp based implementation of the mechanical system. Once the control plant is implemented, the design of a PID controller is undertaken followed by simulation and op-amp based implementation of the closed-loop feedback control system. In this experiment, students are able to independently control the values of proportional, integral, and derivative gains and study their impact on steady state and dynamic performance of a control system. Dynamic performance characterization focuses on damping ratio, overshoot, frequency of oscillation, time constant, rise time, and settling time. Once the effects of individual gains are established, students can study various controller configurations (P, PD, PI, and PID) and eventually design an optimal PID controller by tuning in the parameters. The strength of this experiment lies in the fact that students are in charge of modeling, designing, simulating, implementing, and testing of this simple yet complete system. This experiment can be implemented very inexpensively (needs a couple of quad op-amp ICs and a few discrete resistors and capacitors) and yet provides an opportunity for students to investigate fundamental concepts relating system dynamics to controller configurations in a clear and concise manner. A strong feature of the proposed experiment is that students have full access to the system, allowing them to investigate controller-plant interaction at a fundamental level. Experiments on PID control is abundantly available in literature, however most of them use expensive plants1-3 such as liquid-level system or motor drive system. However a major goal of the experiment presented herein is to develop a pedagogically sound controls experiment that is inexpensive and easily accessible by others. Even though the standard textbooks4-8 cover analysis and design of PID controllers in detail, engineering technology students often find the textbook presentations to be highly mathematical in nature. This impedes students’ ability to achieve a clear and concise understanding of the role each of the parameters (KP, KI, and KD) play in designing a PID controller in relation to steady state and dynamic performance of a system. This observation is applicable to teaching controls to non-electrical engineering students9 as well. Engineering Technology students are generally interested in developing an intuitive understanding, grounded in theory, of PID controller’s functionality. And the expectation is that this understanding should lead them to develop a skillset easily transferable to designing PID controllers in practical systems. This need analysis led to the development of this inexpensive but highly effective mass-spring-damper control system experiment covering analysis, modeling, design, simulation, laboratory implementation, and testing of a PID-controller based feedback control system. The following sections present student outcomes for the proposed experiment, model development for the plant (mass-spring-damper system), step response simulation using MATLAB/SIMULINK, op-amp based plant implementation, PID controller design and simulation, op-amp based closed loop system implementation, and testing. Student outcomes assessment data for the laboratory experiment are also presented along with plans for further improvement to the experiment. Student outcomes for the proposed experiment After conducting the proposed control system experiment, students will develop: • an improved understanding of various controller configurations (P/PD/PI/PID), • an improved ability to design PID controllers for the end-of-semester course project, • an ability to identify which gains (KP, KI, and KD) to be increased and which gains to be decreased in a controller to improve system response, and • an ability to prototype and test an op-amp based system model, given the system transfer function. Open-loop mass-spring-damper system A mass-spring-damper mechanical system10 excited by an external force (f) is shown in Figure 1. This second-order system can be mathematically modeled as a position (x) control system with object mass (m), viscous friction coefficient (b), and spring constant (k) as parameters. Based on a free-body diagram, the system differential equation is expressed in (1). Using Laplace Transform, this time-domain equation leads to a position-to-external force transfer function given by equation (2). Once the transfer function is derived, a set of parameter values (m = 0.1 kg, b = 1 N*s/m and k = 2 N/m) is used to obtain the system transfer function shown in (3).