Modular Dynamic Modeling for Precision Motion Control of Hydraulic Manipulator

Multi-degree-of-freedom hydraulic manipulator is a highly nonlinear and time-varying system with hybrid serial-parallel structures, which makes it difficult to achieve precision motion control. In this study, a modular dynamic modeling (MDM) method is proposed, based on which a model-based motion controller is designed combined with parameter identification. A hybrid serial-parallel hydraulic manipulator includes rotating segments of rigid linkages connected by passive rotating joints driven by a linear hydraulic actuator. Compared with Virtual Decomposition Control (VDC), this paper treats the parallel structure as a whole analytically, avoiding excessive virtual decomposition of the hydraulic manipulator and requiring fewer model parameters. To enhance the control performance based on an accurate hydraulic manipulator dynamics model, this paper replaces the model parameters obtained from the CAD (Computer Aided Design) software with parameter identification. The experimental results show that the proposed controller can effectively improve position-tracking performance in Cartesian space through parameter identification.