Hybrid AI-based Control Strategy for a Full-Vehicle Semi-Active Suspension System Equipped with MR Dampers

This paper’s goal is to design a hybrid AI-based controller for a MR-damped semi-active suspension system using a seven-degree-of-freedom model to enhance ride comfort by mitigating the acceleration exerted on the passenger. A fuzzy-logic control strategy is employed to determine the desired force required to be applied to the vehicle body. Also, a recurrent neural network was developed to predict the input voltage of the damper required to generate its actual force. The fuzzy controller rules are designed based on minimizing the vehicle body acceleration. The neural network is trained using the synthetic data produced from the modified Bouc-Wen model. The performance of the proposed hybrid AI-based method was evaluated by applying it to a full suspension system model incorporating an MR damper model with 7-DOF riding on a bumpy road. Based on various simulations, the effectiveness of the proposed strategy is proved by calculating RMS values of the vertical body acceleration, angular acceleration, and the vertical displacement of body center mass.