MPMC-frame: Multiplatform migration control framework for manipulator control
Guoyu Zuo,Jiyong Zhou,Lu Liu,Daoxiong Gong
DOI: https://doi.org/10.1016/j.conengprac.2024.105844
IF: 4.057
2024-01-19
Control Engineering Practice
Abstract:Existing robot control models suffer from poor generalization performance due to varied tasks between manipulators, configuration differences, and physical limits of motion. To address this problem, a multiplatform migration control framework (MPMC-frame) based on a constrained dynamics model is proposed in this paper. First, a model of the robotic manipulator dynamics with configuration adjustment (CA) is constructed based on a neural network to unify the physical joint data of different manipulators. Second, the model-based controller that can be integrated into the framework is designed, and the constraint on the upper bound on the error of uncertain parameters in the control law to guarantee the control robustness and accuracy of the controller for different manipulator platforms. Final, the generic potential function is designed based on multiplatform task requirements, and a task parameter table is constructed to improve the joint motion control performance of MPMC-frame. The feasibility of the method proposed in this paper are verified through simulations and experiments based on the ABB_irb120 and AUBO_i5 robotic manipulators.
automation & control systems,engineering, electrical & electronic