Neural Network based Vibration Control for Vehicle Active Suspension System

Abstract

This paper presents a novel Neural Network (NN) based vibration control of a Vehicle Active Suspension System (VASS) when subjected to road disturbance for enhancing the travelling comfort to the passengers. The simulation for the vibration control of VASS with Proportional Integral and Derivative (PID) controller is used for training the NN. The nonlinearities of the system parameters can be effectively handled by the NN and also it can deal with unfocused data by considering the prejudiced phenomena such as logical reasoning and perception beyond its domain. The idea of this work is to design, simulate and compare the performance of NN based VASS with that of the uncontrolled suspension system (passive) and the VASS controlled with PID controller. The Root Mean Square (RMS) value of body acceleration as the performance index for genetic optimization, the simulation is carried out using MATLAB/SIMULINK software. Simulation results such as sprung mass displacement, body acceleration, suspension deflection, tyre deflection and Power Spectral Density (PSD) of body acceleration shows the effectiveness of NN in suppression of the vibration of vehicle body compared to passive system, PID based VASS and optimized PID controller.