Vibration Control of Active Suspension System with Mass Uncertainty Based on RBF Neural Network

This paper focuses on the neural network vibration damping control problem of vehicle active suspension system with mass uncertainty under random road disturbances. Its main contribution is that a radial basis function (RBF) neural network controller is proposed to ensure the stability of the resulting closed-loop vehicle active suspension system. Firstly, an uncertain nonlinear active suspension system model is established using the Takagi-Sugeno (T-S) fuzzy model to provide samples for training the neural network observer. Secondly, the optimal vibration control force is derived from minimum principle to offset the inevitable vibrations. Finally, a neural network-based observation system directly derives the controller for improving the damping performance of the suspension system under random road disturbances. By analyzing performance requirements for vehicle active suspension under different simulation scenarios, simulation results demonstrate that the neural network controller designed in this paper can offset the vibration of the vehicle active suspension and show better performance even under complex road conditions.