We propose an estimation method of modeling errors for robot manipulators using a genetic algorithm. First, we compensate nonlinear terms that can be determined by a nominal model of robot manipulators. Then we construct alinearized model. However, this linearized model includes modeling errors in general. The estimation algorithm of modeling errors consists of the following two processes: one is a disturbance observer process (Process A) using an inverse model of a nominal model and the other is genetic algorithm process(Process B) using the results obtained by Process A. A problem of Process Bis the tuning method of parameters which are crossover and mutation probabilities of the genetic algorithm. We mention the tuning method of these parameters. Finally, the effectivness of the proposed method is confirmed by numerical simulation results and from experimental results of a 2 degree-of-freedom manipulator.
展开▼
