Neural network model reference control for semi-active hydro-pneumatic suspension

To improve riding comfort of vehicles, a neural network control strategy with sky-hook as reference model is put forward to deal with the non-linear characteristics of hydro-pneumatic suspension system. On the basis of 2-dof non-linear hydro-pneumatic suspension model, the neural control system's structure was analyzed, and the neural network identifier and controller were designed. Taking the D-class road profile as random road input, through simulation, the performance of the control system was studied with full-load and non-load separately. The result shows that neural network model reference control strategy can effectively decrease the vibration of vehicle body, improve the ride comfort ability and have a good adaptability to the parameter change of the controlled object.