Fuzzy Neural Network Control for Vehicle Stability Utilizing Magnetorheological Suspension System

This paper presents the stability of a passenger vehicle system featuring magnetorheological (MR) suspensions. As a first step, MR damper is devised and its field-dependent damping force is experimentally evaluated. A full-car model equipped with the MR damper is established by considering roll motion which is directly related to the vehicle stability. A fuzzy neural network controller (FNNC) incorporated with the self-learn knowledge is then formulated in order to improve vehicle stability. In addition, in order to eliminate adverse effect of the system coupling, nonlinearity and time delay a correction component for updating the weighting matrix and adjusting controller outputs is designed. Both computer simulation and road test are undertaken in order to demonstrate the effectiveness of the proposed control method. The control results obtained in this work indicate that the proposed control scheme with the MR suspension can considerably improve the stability of vehicles with high performance in vibration isolation.