Novel adaptive sliding-mode control of digital hydraulic systems with nonlinear flow prediction and friction identification

Digital hydraulics has emerged as a novel technology widely utilized in engineering equipment, heavy-duty manipulators, and new energy vehicles. However, the high-frequency discrete fluid generated by high-speed on/off valves (HSVs) exacerbates the nonlinear characteristics of digital hydraulic systems (DHSs), thereby limiting control accuracy during fluid transmission. To address this issue, a model-based adaptive sliding-mode control method (ASMC) is proposed, which incorporates two soft measurement methods that integrate friction identification for the DHS with nonlinear flow prediction for the HSV to accurately describe the kinetic model. Subsequently, the coupling parameters in the Stribeck friction model are precisely identified using the particle swarm optimization-least squares algorithm, replacing previous empirical values. Additionally, a high-precision output flow prediction model for the HSV is constructed utilizing a back propagation neural network to address the drawbacks associated with mechanical inertia in the flowmeter. A second-order integral sliding-mode surface is designed to eliminate steady-state error. By incorporating a boundary layer saturation function, the error jitter can be effectively suppressed, allowing the DHS to converge rapidly to a quasi-sliding mode. Furthermore, the stability of the controlled system is validated by the Lyapunov theory. Results indicate that ASMC significantly enhances the dynamic-static performance of the DHS compared to the traditional integral sliding-mode control method, which overlooks the nonlinear behaviors of output flow and friction force. The response characteristic’s setting time is dramatically reduced from 0.86 s to 0.36 s, while the maximum average steady-state error under various loads greatly decreases from 112.4 μm to 23.4 μm. Therefore, the proposed ASMC with the two soft measurement methods presents an innovative solution for the high-precision motion control of the DHS and holds significant engineering application value.