A novel dynamic model parameters identification method for robotic manipulator based on the feasible physical constraints

The dynamic model of robots plays a crucial role in control tasks. The traditional Coulomb-viscous friction model inadequately describes friction torque, and the conventional identification strategy is ineffective in accurately identifying dynamic models with full physical parameters while meeting physically feasible constraints. In this paper, we propose a two-step strategy for identifying dynamic model parameters of manipulators based on nonlinear friction models. We establish the manipulator dynamic model using an improved Stribeck friction model. Firstly, the manipulator and friction force are separated, maintaining the nonlinear characteristics of the friction force while linearizing the manipulator. Secondly, the two components are optimized separately, incorporating physical feasibility constraints throughout the process. To evaluate the proposed method, a series of comparison experiments are conducted on the xMate Pro 3 manipulator plant. Experimental results demonstrate that the two-step identification strategy significantly improves identification accuracy. Meanwhile, the improved Stribeck friction model improves the comprehensive accuracy of parameter identification by 46%.