ADAPTIVE CONTROL OF A MANIPULATOR AND ITS MICROPROCESSOR IMPLEMENTATIONS

The main area of this research is to investigate the approach of adaptive control technique (self-tuning type) to the control of a manipulator and the implementation of the algorithm on a microprocessor system. A linearized model of the dynamic equations of the manipulator is obtained by the perturbation method about a desired trajectory. The parameter values of the linearized model along a desired path are calculated to demonstrate the variability of the parameters when the manipulator is moving. Based on the input/output data of a manipulator, the detailed dynamic model of a manipulator is represented by a set of autoregressive difference equations. The parameters of the model are calculated by an on-line recursive algorithm based on the least squares error criterion. The algorithm tracks the values of the model parameters which change slowly along the path, as is demonstrated by experiments. An adaptive controller is designed on the basis of the autoregressive model using estimated parameters and a selected performance criterion. The optimal inputs based on this model are calculated on-line. Simulation studies are presented to demonstrate the applicability of this adaptive control scheme. The adaptive control scheme of self-tuning type has been successfully implemented on a microprocessor system for three joints of a manipulator. The results are satisfactory considering the limited capabilities of the microprocessor system and other equipment