Non-Linear Predictive Control of Multi-Input Multi-Output Vehicle Suspension System

Abstract

A new predictive control approach is employed to control a non-linear 4 degrees of freedom vehicle suspension model including pitch and bounce motions. The front and rear suspension actuator forces are considered as control inputs. These are determined by minimizing a performance index defined as a weighted combination of conflicting objectives, namely ride comfort and road holding. In this way, the suspension deflection and the control forces should be restricted to admissible practical ranges. In this paper, an optimization process is used to develop a multi-input multi-output (MIMO) non-linear control law in a closed-form which is mathematically tractable and suitable for implementation. To investigate special features of the derived control law, the closed-loop system dynamics are first evaluated analytically and then the corresponding simulation results are presented. The results show that a compromise between mentioned objectives and control energy, even in the presence of uncertainties, can be easily made by regulating the control weighting factors.