PID Parameters Prediction Using Neural Network for A Linear Quarter Car Suspension Control

Abstract

Providing control for suspension systems in vehicles is an enhancing factor for comfort and safety. With the improvement of control conditions, it is possible to design a cost-efficient controller which will maintain optimum comfort within harsher environmental conditions. The aim of this study is to design an adaptive PID controller with a predictive neural network model, which will be referred as NPID (NeuralPID), to control a suspension system. For this purpose, a NN (Neural Network) model is designed to produce outputs for PID’s Proportional (P) parameter to provide optimum responses for different road inputs. Also, reliability of the system outputs, which is using adaptive Proportional parameter, is tested. PID parameters for linear quarter vehicle model are decided through Zeigler-Nichols method. An ideal PID model, where Integral (I) and Derivative (D) parameters are bound to Proportional parameter, is used in the system. When the outputs of different controlled and not controlled systems, which are free, PID and NPID, are compared; it has been seen that NPID outputs are more convenient. In addition, it is possible to design controllers, with adaptively adjusting P parameter, which are operating cost-effective.