关键词: 运动学参数标定, 蜣螂优化算法, 精英反向学习, 鱼鹰探索行为, 随机扰动机制

Abstract: Aiming at the DBO algorithms imbalance between global exploration and local exploitation capabilities and low solution accuracy in the calibrating processes for kinematics parameters of industrial robots, a multi-strategy fusion(MSFDBO) algorithm was presented based on local product of exponential(LPOE) kinematics model. Firstly, a kinematics parameter identification model was established based on the LPOE model. Secondly, Piecewise chaotic mapping and elite inverse learning strategy were used for population initialization to obtain a more uniformly distributed population, incorporating the exploration behavior of the osprey to enhance the global exploration ability of the DBO algorithm, and expanding the search range through the stochastic perturbation mechanism to reduce the possibility of the DBO algorithm falling into a local optimum. To test the performance of the algorithm, the search performance of the MSFDBO algorithm was experimentally evaluated using 12 benchmark test functions. The results show that the algorithm performs well in terms of optimization. The compensation of kinematic parameters was identifed and verified for four T6A-19 industrial robots. The experimental results show that the mean absolute position errors are reduced by an average of 85.47% and the root mean square errors are reduced by an average of 83.92%.

Key words:  , kinematics parameter calibration, dung beetle optimization(DBO) algorithm, elite opposition-based learning, osprey exploratory behavior, stochastic perturbation mechanism