RBF Neural Network-based Adaptive PID Controller for Active Suspension

Abstract

Active suspension plays a pivotal role in modern vehicles. In this paper, an adaptive PID controller of active suspension systems based on RBF neural network (RBF-NN) is developed. A quarter-car suspension system with two degrees of freedom is demonstrated. The values of proportional, integral, and derivate components are obtained by using Ziegler-Nichols(Z-N) tuning method and RBF-NN methods. The suspension system is perturbed using the sine function. Simulated in the Simulink environment is the quarter-car model. Passive suspension systems, adaptive PID controller utilizing the Z-N tuning approach, and adaptive PID based on the RBF-NN method for active suspension systems are compared. The active suspension with PID control based on the RBF-NN outperformed the active suspension with PID control utilizing the Z-N tuning approach and passive suspension, according to simulation data. The comparison demonstrates the proposed control method’s superior features