SUSPENSION SYSTEM'S VIBRATION CONTROL: A PARAMETRIC UNCERTAINTIES PROBLEM IN CONTROL ENGINEERING

Abstract

In this dissertation work, QFT and H-infinity controllers were designed for a vehicle suspension system and their performance is compared and also with a standard PID controller. A suspension system works in uncertainties, which implies variation in plant parameters. Hence in designing controller for this system, sensitivity, which can be defined as variation in response affected due to change in parameters, and robustness analysis are primary issues. The performance objective for the controller design is to keep the error between the controlled output and the set-point as small as possible and sensitivity function reshaping is the main design tool utilized. In the Quantitative Feedback Theory (QFT) controller, which is a classical approach to design a controller suited for systems having large uncertainties by using feedback of measurable plant outputs to generate an acceptable response from a system with disturbance signals and plant uncertainties, measured displacement of the suspension system is used as a feedback. The QFT controller is based on reshaping of the loop-transmission function (product of plant and controller transferS function) on which sensitivity of the system depends. The QFT controller is implemented in MATLAB. The H-infinity (H~) controller, which is also a frequency response approach used to design a robust controller that can reshape the sensitivity function directly to achieve the desire performance objectives, is done by selection of suitable weight function. The controller is also implemented in MATLAB and it realizes the sensitivity reshaping. The QFT and H-infinity controllers outperform the trial and error method of PID. Simulation results have been included for all.