Digital pendulum system: Genetic fuzzy-based online tuning of PID controller

Abstract

Main aim of this research is to develop a robust controller for an inverted pendulum system. Performance of classical PID controller is found to be effective in this regard. However, effectiveness of PID controller depends on its three gain values that require proper tuning. Two different tuning methods have been adopted in this study. In the first approach, frequency response-based Zeigler Nichols PID tuning has been considered. In the second approach, a Fuzzy Logic (FL)-based tuning of PID controller gains has been implemented. Moreover, performance of FL-based tuner has been optimized using a binary coded genetic algorithm. It is observed that control performance of FL-based method is substantially better compared to the other method. It may be due to the fact that FL-based method is not taking into account the nonlinearities and plant uncertainties present in the model explicitly. Both the simulation and experimental analysis have been carried out in MatLab Simulink environment.