A new method of adapting the parameters of PID controllers

We propose a new adaptation mechanism for the parameters of PID controllers using fuzzy logic. The fuzzy logic adaptation mechanism consists of two separate units. The first unit permanently describes the dynamic behaviour of the plant by using graphic linguistic variables like the phase angle. On the basis of those linguistic variables, the second unit calculated adaptation factors for the different controller parameters. The proposed fuzzy logic adaptation mechanism was tested using the six-degree-of-freedom-robot RV15, manufactured by Reis Robotics. The measurements show that the Cartesian position error of the tool center point (TCP) can be significantly reduced by 50%. The influence of different loads on the Cartesian position is decreased as well.