Nonlinear dynamic compensation for a compact high-sensitivity ball double-screw joint actuator in legged robots

• A compact ball double-screw hydraulic actuator achieves high sensitivity and over 70% volume reduction. • A nonlinear temperature–current drift model with adaptive fuzzy compensation improves control precision. • Simulations and experiments confirm ±0.03° positioning and <0.5° tracking errors across 10–60 °C. The ball double-screw robot joint actuator is a compact, high-energy-density rotary drive device featuring rapid response and minimal frictional resistance. Unlike traditional hydraulic actuators, which are often bulky and slow to respond, this novel actuator significantly reduces size through modular servo control design while delivering motion performance that is comparable to electric joints. This makes it especially promising for compact, heavy-load robotic applications that demand high-precision motion control. However, due to inherent nonlinearities and environmental uncertainties, precise control remains challenging. To address this, an adaptive fuzzy control strategy with real-time drift-current compensation is proposed. A nonlinear temperature–current drift model, developed from experimental data and theoretical analysis, is integrated with a moving-window mechanism for dynamic compensation of temperature-induced drift and other disturbances. Experimental and simulation results over a 10–60 °C temperature range demonstrate marked improvements in positioning accuracy and dynamic tracking precision, with errors maintained within 0.03° and 0.5°, respectively. The proposed method effectively enhances the control performance of compact hydraulic actuators, providing a robust solution for high-precision control in demanding robotic systems.