Genetic optimization of a fuzzy active suspension system based on human sensitivity to the transmitted vibrations

Abstract

Optimization of a vehicle fuzzy active suspension (AS) controller was previously performed on the basis of the amplitude of transmitted vibrations to the body. However, ride comfort depends strongly on the human sensitivity, which is a function of not only the amplitude but also the frequency contents of the transmitted vibrations. In this paper, genetic optimization of a fuzzy AS system based on the human sensitivity to the transmitted vibrations is presented. For this purpose, a fuzzy logic controller (FLC) is initially proposed for the AS system control. The FLC parameters are then tuned using a genetic algorithm (GA). The tuning process is first formulated as a single-objective optimization problem based on the human sensitivity and conventional r.m.s. amplitude criteria separately. The simulation results reveal that the optimization of a fuzzy AS based on the common r.m.s. amplitude criterion not only does not result in the optimal ride index, but also causes a considerable increase in the energy consumption. Moreover, in order to accommodate the conflicting characteristics of the AS system, the FLC parameters are tuned on the basis of a multi-objective fitness function incorporating human sensitivity, suspension travel, and energy consumption. The simulation results prove the effectiveness of the optimized FLC in hitting the simultaneous targets of ride comfort improvement as well as suspension travel and energy consumption reduction.