Error Identification of Robot Geometric Parameters Based on GA-SA Algorithm
The geometric parameter error has the greatest influence on the absolute positioning ac-curacy of the robot end,and the geometric error parameter identification is a high-dimensional nonlin-ear problem,which is difficult to solve,so it is necessary to establish a simple and efficient identification algorithm,and this paper proposes the Genetic Simulated Annealing Algorithm(GA-SA)for the iden-tification of the robot geometric parameter error.With the goal of minimizing the robot end position er-ror,the genetic simulated annealing algorithm is used to identify the robot geometric parameter error.1100 iterations of ABB IRB120 are used as an example,and the genetic algorithm falls into the local op-timum in 200 generations,and the final fitness after simulated annealing is 0.0914.The results of the error compensation show that the robot end position error along the axial direction is reduced by 88.05％,81.73％,83.72％,and the attitude error is reduced by 93.92％,83.64％,83.44％,respectively,which proves that the genetic simulated annealing algorithm can effectively identify the robot geometric parameter errors and improve the robot end position accuracy after error compensation.
万方数据
维普
