The electro-hydraulic system is widely used in industrial production due to its high power-to-weight ratio, but the heavy-duty characteristics make the electro-hydraulic system subject to large disturbance force even if the actuator moves slightly, especially in mobile machines and multi-actuators system. Therefore, a position and velocity constraints method based on barrier Lyapunov function is proposed to guarantee the tracking error limited in a strict range to avoid the large disturbance force. Besides, the external disturbance, parameters uncertainty and modeling errors in the asymmetric cylinder electro-hydraulic systems affect the accuracy of position tracking seriously. So a high-gain disturbance observer is designed to estimate the lumped disturbance of the system, which can avoid amplification of the noise during the states measurement. In addition, dynamic surface control based on backstepping method is adopted to avoid the derivative explosion phenomenon when calculating the derivatives of virtual control inputs, which reduces the computational complexity of the system significantly. To verify the effectiveness of the proposed controller, proportional-integral controller and adaptive controller are designed to be compared with the high-gain disturbance observer–based dynamic surface controller with the backstepping method, and the comparison results show that the proposed controller has a more precise trajectory tracking performance.
Adaptive neural network dynamic surface control for musculoskeletal robots
