Intelligent active force control of a vehicle suspension system

Abstract

Active suspension control aims to suppress the undesirable vibration and other loading effects and should provide improvements in term of passenger comfort. This study deals with the design and implementation of robust active force control (AFC)-based schemes that incorporates artificial intelligence techniques plus a number of feedback control strategies applied to a vehicle suspension system. The overall proposed control system essentially comprises four feedback control loops, namely, an innermost loop for force tracking of the pneumatic actuator using a proportional-integral controller, two intermediate loops applying the skyhook and AFC strategy for the compensation of the disturbances and an outermost loop for the computation of the desired force for the actuator using a proportional-integral-derivative controller. Adaptive neural network and adaptive fuzzy were proposed and employed to compute the inverse dynamics of the nonlinear pneumatic actuator and estimated mass of the system within the AFC loop. The integration of all the interrelated elements leads to the formation of two main proposed schemes known as the Skyhook Adaptive Fuzzy Active Force Control and Skyhook Adaptive Neuro Active Force Control. The suspension system was modelled based on a two degree-of-freedom quarter car configuration. A number of road profiles were also modelled as the main disturbance elements to evaluate the system robustness and vehicle dynamic performance related to ride comfort. Simulation results both in time and frequency domains demonstrate the effectiveness of the proposed AFC-based schemes in countering the disturbances and other loading conditions. The schemes show evidence of at least 33.9% improvement in performance over the passive suspension. This is complemented by an experimental study on a developed full scale quarter car suspension test rig which shows a very good agreement with the simulation counterpart.