Exploiting Fractional Order PID Controller Methods in Improving the Performance of Integer Order PID Controllers: A GA Based Approach

The paper is divided into three parts The first part gives a brief introduction to the overall paper, to fractional order PID ((PID mu)-D-lambda) controllers and to Genetic Algorithm (GA). In the second part, first it has been studied how the performance of an integer order PID controller deteriorates when implemented with lossy capacitors in its analog realization Thereafter it has been shown that the lossy capacitors can be effectively modeled by fractional order terms Then, a novel GA based method has been proposed to tune the controller parameters such that the original performance is retained even though realized with the same lossy capacitors Simulation results have been presented to validate the usefulness of the method Some Ziegler-Nichols type tuning rules for design of fractional order PID controllers have been proposed in the literature [11] In the third part, a novel GA based method has been proposed which shows how equivalent integer order PID controllers can be obtained which will give performance level similar to those of the fractional order PID controllers thereby removing the complexity involved in the implementation of the latter It has been shown with extensive simulation results that the equivalent integer order PID controllers more or less retain the robustness and iso-damping properties of the original fractional order PID controllers Simulation results also show that the equivalent integer order PID controllers are more robust than the normal Ziegler-Nichols tuned PID controllers