Application of a self-organizing fuzzy control to the joint control of a Puma-760 robot

The control performance of the present self-organizing control (SOC) algorithm is often limited due to quantization of process variables and large amount of computation time required within small sampling time. To improve these limitations this paper presents a combined control structure in which the conventional P-control law is used in parallel with a modified fuzzy SOC algorithm. To run the algorithm fast, the inference mechanism of the fuzzy rules is simplified by modifying the SOC algorithm. Addition of the P-control law is made in the feedforward path to guarantee stability at initial stage before the rules are completely created. To test the performance of the proposed control scheme and structure, a series of experiments were performed using a PUMA-760 robot. The experimental results show that the proposed structure can be effectively applied to the control of robot manipulators when implemented to the personal computer