摘要
机器人的惯性参数辨识是基于动力学模型控制器设计的基础,除了需要辨识机器人自身的惯性参数外,还需要快速准确地辨识负载的惯性参数。针对需要更换负载的工作场景,提出了一种快速辨识负载的激励轨迹,该轨迹不需要经过优化,即可在短时间内覆盖更多的机器人运动状态,从而获得更为全面的采样参数集,提高了辨识效率的同时增强了辨识模型的泛化能力。实验表明,辨识的结果与负载模型的CAD参数符合,机器人关节理论力矩能较好地拟合实际驱动力矩,验证了负载参数的正确性。
Inertial parameter identification of robot is the basis of controller design based on dynamic model.In addition to identifying the inertial parameters of robot itself,it is also necessary to quickly and accurately identify the dynamic inertial parameters of load.For the working scene where the load needs to be replaced,an excitation trajectory that can quickly identify the load is proposed.The trajectory can cover more robot motion states in a short time without optimization,so as to obtain a more comprehensive set of sampling parameters,which improves the identification efficiency and enhances the generalization ability of the identification model.The experimental results show that the identification results are consistent with the CAD parameters of the load model,and the theoretical torque of the robot joint can better fit the actual driving torque,which verifies the correctness of the load parameters.
作者
钱鸿巍
李成刚
杜兆才
李檬
储亚东
QIAN Hongwei;LI Chenggang;DU Zhaocai;LI Meng;CHU Yadong(College of Mechanical and Electrical Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China;Key Laboratory of Digital Manufacturing Aviation Science and Technology,China Aerospace Manufacturing Technology Research Institute,Beijing 100024,China)
出处
《机械与电子》
2023年第4期65-70,共6页
Machinery & Electronics
基金
南京航空航天大学研究生科研与实践创新项目(xcxjh20210501)
中国航空科学基金项目(201916052001)。
关键词
机器人
动力学
参数辨识
激励轨迹
robot
dynamics
parameter identification
excitation trajectory
