Dynamics parameter identification of industrial robots based on immune algorithm

Dynamics parameter identification is a key factor and a difficult area in the development of robot motion control technology. In order to obtain accurate dynamics parameters, an integral identification method combined with the immune algorithm is studied and experimented with. Firstly, the dynamics model of the industrial robot is derived from the Newton-Euler iteration method. Then, taking the condition number of the observation matrix as the optimization index, an ideal excitation trajectory is optimized by the immune algorithm. Butterworth zero-phase low-pass filter and the weighted least square method are used to solve the minimum parameter set of dynamics. A new validation trajectory is designed using the fmincon toolbox to cross-validate the identified dynamics model. Finally, the above identification algorithm is tested with Inexbot LG-607L robot in China as the experimental object. The experimental results show that the immune algorithm can calculate the excitation trajectory with a lower matrix condition number than the fmincon method and the genetic algorithm. And the parameter identification effect of the integral method is accurate.