Changzhi Chen, Lin Jia, Teng Chen, Guoteng Zhang, Xuewen Rong
Hydraulic manipulators are widely used in heavy-load automation, yet nonlinear coupling and external disturbances hinder high-precision trajectory tracking. Notably, it remains challenging to guarantee transient error bounds while maintaining near-optimal performance. This paper proposes a disturbance-compensated prescribed-performance optimal backstepping control scheme for a multi-DOF hydraulic manipulator. Within the backstepping framework, an error transformation enforces prescribed performance bounds on transient and steady-state tracking errors. A weighted cost function is constructed and the Hamilton-Jacobi-Bellman equation is derived to integrate optimality into the backstepping design. An actor-critic network approximates the value function and yields near-optimal control compensation. A disturbance observer further estimates and rejects lumped disturbances to enhance robustness. Lyapunov analysis establishes semi-global uniform ultimate boundedness of the closed-loop signals. Finally, experiments validate the proposed method's effectiveness.