Variable-Universe Fuzzy Adaptive Sliding-Mode Control of the Hydraulic Manipulator of a Salix psammophila Coppicing Machine

Coppicing operations for Salix psammophila are commonly conducted on markedly uneven terrain and under highly variable cutting loads. Under these conditions, the hydraulic manipulator is susceptible to hydraulic parameter variations and external disturbances. These conditions make accurate trajectory tracking difficult and may affect the operational requirement of maintaining the remaining stump height below 5 cm. Although variable-universe fuzzy control and adaptive fuzzy sliding-mode control have been investigated in nonlinear and electrohydraulic systems, their application-oriented integration and full-scale validation for hydraulic manipulators under terrain- and load-varying coppicing conditions remain limited. To address this gap, a variable-universe fuzzy adaptive sliding-mode controller is developed and experimentally validated for the hydraulic manipulator of a Salix psammophila coppicing machine. The contribution lies in the application-specific integration of adaptive uncertainty compensation, error-dependent variable-universe fuzzy regulation, and full-scale experimental validation. An adaptive law is employed to estimate unknown system parameters and equivalent disturbances online. Fuzzy regulation adjusts the switching control to mitigate chattering, while the variable-universe mechanism improves the response under large errors and the accuracy under small errors. Closed-loop stability is established using Lyapunov theory. MATLAB/Simulink R2024b simulations incorporating prescribed hydraulic-parameter variations and external-load disturbances, together with full-scale experiments, were conducted to validate the proposed method. The simulation results show that VFASMC improves trajectory-tracking accuracy relative to SMC and fixed-universe FASMC and achieves lower RMSEs than ASMC while maintaining smoother control inputs. In the full-scale experiments, the mean RMSE averaged over all tested hydraulic cylinders and terrain conditions was reduced by 80.1% relative to FASMC and by 67.4% relative to ASMC. Millimeter-level tracking accuracy was achieved, and the residual stump height was maintained below 5 cm. These results demonstrate the practical feasibility of the proposed controller under the investigated coppicing conditions.