Fuzzy control of an industrial robot in transputer environment

The application of the fuzzy logic controller (FLC) to the control of an industrial robot and the experimental results are described. The performance of this controller was investigated by feeding step inputs to several joints at the same time. The results are compared with those resulting from a conventional proportional-derivative (PD) controller. One disadvantage of the FLC is the difficulty of designing an optimal rule base. To overcome this problem a self-organizing controller was implemented and tested. Due to the demand for a small sample time, a multiprocessor environment was used in experiments. The calculation of the control for the individual joints was shared by the network of transmitters. In this network processes can run effectively in parallel.<<ETX>>