A comparative study of an optimal controller and a fuzzy logic controller for a multi-input multi-output system

This paper deals with a comparison between the performance of a dynamic system when controlled according to a fuzzy logic control policy and when governed by an optimal controller incorporating integral action for disturbance rejection. The system under study has two inputs and two outputs. For design of an optimal controller, however, an integral action is added to the controller which leads to an augmented system of fourth order. This system is initially unstable; it is linear but is subject to certain intrinsic and input constraints. This makes the system quite interesting from a control study point of view. In particular the system in question is a vehicle equipped with front and rear steering wheels. The control problem for path following of such a vehicle can be based on the kinematics only at lower speeds. The system can be treated as a regulator problem when following a straight path, but for following curves a tracking problem is involved. The behaviour of the controlled system is simulated under the influence of each of the two controllers, and various matters associated with design and effects of controllers are compared.