GMS model-compensated adaptive model predictive control of electro-hydraulic servo system

Electro-hydraulic servo tracking is degraded by nonlinear friction, particularly at low speed. This study proposes a GMS model-compensated adaptive model predictive controller. The identified GMS model describes presliding and sliding friction and provides feedforward compensation; adaptive PD improves transient response, and MPC handles constrained tracking. Simulations with ramp and sinusoidal inputs show smaller low-speed fluctuations and tracking errors than PID control. Over 10–18 Hz, the maximum phase and amplitude errors are 5.3839° and 1.4880%. At 18 Hz, they decrease from 9.9985° to 20.2756% under PID control to 3.7346° and 0.7344%, respectively. Thus, the proposed controller satisfies the “double-ten index” over the evaluated range and shows potential for electro-hydraulic servo systems with nonlinear friction.