Model Predictive Control for Active Suspensions: Controller Design and Experimental Study

Abstract

Abstract This paper discusses the design of a Model Predictive Controller for a vehicle active suspension system and presents the experimental results obtained in the evaluation of the controller performance. The Model Predictive Control methodology described here, utilizes the road profile ahead of the vehicle as preview information and also takes into explicit consideration the physical constraints on the suspension travel. The advantages in the use of preview information in active suspension control have been studied by several researchers. The explicit consideration of the physical limits on the suspension travel becomes an important factor under rough road conditions such as the ones encountered by off-road vehicles. A computationally efficient version of the predictive controller was developed so as to be implementable on-line. The experimental evaluation of the controller was performed on the UC Berkeley Active Suspension Test Rig.