Design a new algorithm (iL-GA) to optimize controller parameters for an automotive suspension system

Abstract

Purpose This paper aims to design a PD controller for an active suspension system to improve the car’s moving smoothness. Design/methodology/approach The controller parameters are optimized by an in-loop genetic algorithm (iL-GA). Unlike previous studies that only used conventional GAs to tune coefficients for the controller, the iL-GA designed in this paper provides outstanding efficiency when determining the optimal value range for the system. The optimal value range of parameters is determined by the in-loop algorithm based on criteria related to systematic errors. The optimal values are then calculated by the GA based on this range instead of an uncertain one. Findings Simulation results show that vehicle body acceleration and displacement values are significantly reduced when using the active suspension system compared to the conventional passive suspension system. The phase difference phenomenon does not occur in the iL-GA situation. In addition, the frequency domain investigation also shows the system’s stability when using iL-GA instead of conventional GA. Originality/value To the best of the authors’ knowledge, this is a new application that provides positive effects to the suspension controller. This algorithm can be applied to tune coefficients for direct controllers in the future.