Genetic algorithms for robot control

The control of nonlinear systems (for example robot manipulators) is a formidable problem. We propose to solve this problem by incorporating a genetic algorithm with an inverse dynamics model to generate a reasonable control signal. The method is tested by using the positioning system (the first three degrees of freedom) of the Stanford and Puma manipulators. The controller is used to generate correction signals to compensate for dynamic parameter variations between the model used and the actual manipulator. The results show that the method is an efficient tool for the control of nonlinear systems