Design an Artificial Neural Network Based Predictive Model for Automotive Applications

Abstract

Passive shock absorbers can be designed for standard load condition, which provide better performance of vibration isolation only for standard load condition. Passive fluid shock absorber gives effective solution for vehicle’s comfort as well ashandling and convertsenergy of vibration into heat using throttling viscous fluid by restricted orifice which can be highly utilized for vehicle suspensions. Research aims to identifying the peak displacement, peak acceleration, power and stiffness for a given frequency and number of holes opens. For this purpose, Artificial Neural Network (ANN) is utilized for identifying the above parameters and the associates training techniques are RP (Resilient Back-propagation), BFGS Quasi-Newton (BFG), SCG (Scaled Conjugate Gradient). The proposed SCG attains minimum error values as 18.74 that is lower that other comparative techniques.