Research on Identification Method of Kinematics for Space Robot

摘要

As the space environment and other factors, the space robot kinematic parameters will be some changes while running in the orbit. In order to improve the positioning accuracy of space robots a kinematics parameters identification method is proposed which combining linear simplification with nonlinear optimization, using a combined identification strategy of least squares algorithm with the PSO algorithm. In the initial phase, the nonlinear model is identified off-line using the (Particle Swarm Optimization) PSO algorithm. While in the follow-up mission, the kinematic parameters do not change much, the least squares method is used for online identification. This method can satisfy the identification accuracy in conditions of large deviations, and to achieve kinematic parameters online identification. Simulation results demonstrate the effectiveness of the method.