Research on suspension system based on genetic algorithm and neural network control

Abstract

In this paper, a five degree of freedom half body vehicle suspension system is developed and the road roughness intensity is modeled as a filtered white noise stochastic process. Genetic algorithm and neural network control are used to control the suspension system. The desired objective is proposed as the minimization of a multi-objective function formed by the combination of not only sprung mass acceleration, pitching acceleration, suspension travel and dynamic load, but also the passenger acceleration. With the aid of software Matlab/Simulink, the simulation model is achieved. Simulation results demonstrate that the proposed active suspension system proves to be effective in the ride comfort and drive stability enhancement of the suspension system. A mechanical dynamic model of the five degree of freedom half body of vehicle suspension system is also simulated and analyzed by using software Adams. © Tang and Guo; Licensee Bentham Open.