Multiple Model Adaptive Control of a Nonlinear Active Vehicle Suspension

Widely varying operating mode dynamics, uncertain parameters and nonlinear behaviour are inherrent characteristics of vehicle suspensions. Mechatronic vehicle suspensions seek to produce improved ride and handling performance by providing active control action that addresses these problems. Inappropriately designed controllers may lead to poor performance or even possible instability. In this work, a multiple model adaptive control scheme is proposed for a non linear mechatronic suspension subjected to variations in sprung mass and road profile input. Proportional-Integral-Derivative (YID) candidate controllers corresponding to four operating mode conditions were optimally designed a priori. Weighted control inputs from these candidate controllers were then used to form the actual control action on the plant. To ensure stability, the control system was designed within the framework of adaptive mixing control. Results from simulation tests showed significantly improved performance for both sprung mass acceleration (corresponding to ride) and tyre force (corresponding to handling) compared to that of passive suspensions. Based on RMS values of sprung mass acceleration and tyre force, the proposed control scheme also produced improved performance over LQG and mu-synthesis controlled active systems.