Neuro model for passive suspension of a light car

Abstract

The system model is neccessery to be determined in control systems engineering which is generally represented in mathematical form. The mathematical model can be utilized for analysing the system’s characteristics or designing the controller parameters of the system. Here, a neuro model for passive suspension system of a light car is proposed. The candidate structure of the neuro model is contructed from non-linear system of passive suspension of a quarter car mathematical model. Weights estimation of neuro model is conducted by applying iterative weighted least square algorithm. Actual input output data of a test car for training process are acquired by driving the test vehicle on an artificial surface of road. An artificial surface of road is a kind of real road surface imitation. Experimental findings show that the proposed model is able to imitate sucessfully the dynamic properties of the passive suspension system of the light car. The model response shows similar trend and has smallest error.