Fuzzy wavelet neural network-based backstepping control for electro-hydraulic servo systems

For electro-hydraulic servo systems (EHSSs), inherent nonlinearities and external disturbance degrades its dynamic tracking accuracy. In this paper, the backstepping technique is applied for the fuzzy wavelet neural network (FWNN) to deal with the difficulties caused by nonlinearities and external disturbance. Specially, wavelet analysis and fuzzy logic are employed to optimize the structure of neural network (NN) during the backstepping recursion process. All parameters of NN weights are tuned online without any prior information. In addition, the robust terms are synthesized into the controller design to improve the closed-system tracing performance. More importantly, the global stability of the EHSSs is strictly guaranteed by the Lyapunov analysis. Simulation comparative results indicate that the proposed control strategy can achieve high tracking accuracy and good robust performance.