Nonlinear adaptive backstepping control of a friction based electro-hydraulic load simulator using chebyshev neural networks

This paper deals with the output torque tracking control of a friction based electro-hydraulic load simulator (FEHLS) with time-varying spring stiffness and bearing friction. Although the FEHLS has no extra torque, the friction of the bearings will cause uncertain friction torque. Besides, due to the special mechanical structure of the FEHLS, the spring stiffness will vary near the zero position of the hydraulic cylinder and also cause uncertain torque. To compensate for the uncertain torque and improve the torque tracking accuracy of the FEHLS, an adaptive backstepping controller based on Chebyshev neural network (CNN) is proposed. First, the nonlinear control model of the FEHLS is established where the uncertain torque is lumped into unknown disturbance. Then, the CNN is used to approximate the lumped unknown distrubance. Based on the CNN approximation, an adaptive backstepping controller is designed. Finally, simulation studies are carried out to show the effectiveness of the proposed controller.