Parameter Design of Model Predictive Control for the Displacement Control of an Axial Piston Pump

The tilt angle control of variable displacement hydraulic pumps is susceptible to changes in dynamic characteristics that depend on operating points and load pressure, and hence high control performance is hard to achieve. In this study, a variable displacement pump without a mechanical feedback mechanism is focused on and its displacement control system is designed using a model predictive control (MPC) method that can consider various design constraints. In particular, the adaptive MPC is introduced for accepting dynamic characteristics change and load pressure. The results clarified that the difference due to the variations in the model order was negligible for the steady-state deviation and settling time; however, the overshoot variation and the average of absolute tracking error could be attenuated using a higher order model for load pressure changes. Furthermore, the settling time and tracking error were considerably reduced by increasing the prediction and control horizons properly.