Facing the field of intelligent manufacturing, aiming at the dexterous operation and intelligent assembly problems of typical connectors, with the help of the perception mechanism of multi-modal information fusion, we study the independent planning of complex operating procedures of robots, complete the modeling and optimization of assembly sequences, and realize the dexterous operation of robots and intelligent autonomous assembly.
1.Preliminary research basis and results
(1) Accurate docking of target objects with multi-degree-of-freedom manipulators a model for the docking of robotic arms to target objects was established, and a flexible controller for the robotic arm that integrated visual and force sense was proposed. Through the fusion of visual and force sense information, the robotic arm was ensured to insert accurately, and the demonstration and verification of the robotic arm docking were achieved. The results were published as the first author in the well-known international conference IEEE-ICRA in the field of robotics (pp.1036-1041, 2014)
(2) Double-arm cooperative linear flexible body intelligent assembly taking the two-arm collaboration to complete the assembly of computer power cords as an example, a bionics-based dual-arm collaboration method for linear flexible body assembly is proposed, and a manipulator for linear flexible body assembly is innovatively designed. Part of the results have been published at the workshop ( Workshop: BIMANUAL MANIPULATION: LEARNING, PLANNING AND CONTROL) of IEEE-ICRA 2021, a well-known international conference in the field of robotics.
2.Industrial application and technical services
(1)Industrial application
1) Intelligent precision assembly of robotic arms for 3C and other industries.
2)Robotic arm-assisted laser fully automatic marking, including marking of workpieces placed in any position in space and marking of linear flexible bodies.
(2)Technical services
1) Customized robot digital twin system (purely developed in Python , does not use third-party software, does not use third-party platforms such as ROS.