Adaptive and nonlinear model-based control of variable displacement pumps for variable loading conditions

Abstract

Energy efficiency remains one of the most important unsolved challenges in hydraulic (robotic) systems, especially in off-highway machines and ambulatory robotic systems where the energy source(s) must be carried on board in limited space. Consequently, this paper proposes an adaptive and nonlinear model-based (NMB) controller for a discharge pressure-controlling variable displacement axial piston pump (VDAPP), to reduce the energy consumption of the hydraulic systems. For the control design, a novel NMB control method is used. The method originates from the virtual decomposition control (VDC) approach but is developed in a novel manner. Similar to VDC, the control design is based on virtually decomposed subsystems' dynamics, rather than centralized system dynamics, while rigorously guaranteeing asymptotic convergence. The dynamic interaction among subsystems is addressed with the designed stabilizing term and stability-preventing term, which eventually cancel each other out when subsystems are coupled. The rationale for this study is to 1) examine the novel control design method in comparison with state-of-the-art controllers for discharge pressure controlling VDAPPs and 2) design a high-performance discharge pressure-tracking controller for a VDAPP that supplies an actively controlled load valve. The latter is to obtain reductions in energy consumption compared to conventional methods. The comparative experiments demonstrate an advancement of the proposed method.