Stabilization of cart inverted pendulum system using double pid controller with genetic algorithm based optimal tuning

Abstract

Since the 1940s, industries have utilized the proportional-integral-derivative controller, or PID Controller, extensively and it is still the most popular. The traditional tuning of PID controllers for different applications require great knowledge and experience to arrive at the optimal values. To mitigate this issue, a plethora of techniques are reported in the literature. In this research, a Genetic Algorithm (GA) based approach is used to optimize the PID Controller parameters for stabilization of a cart inverted pendulum system. The present work proposes a double PID control scheme for cart position and pendulum angle and aims to tune both the PID controllers using GA. In this regard, first, it is necessary to acquire the stable area for the stabilization problem, which is achieved here by using a trial and error approach involving GA. Finally, the optimal Kp, Ki and Kd values for both the PID controllers are obtained inside the stable area by using GA based technique. The thesis then compares the results obtained using different variations of GA parameters and decides the optimal one. It is further compared with results involving single-PID control schemes tuned with different traditional control techniques. It shows that the proposed GA based double PID control scheme yields superior results in terms of transient performance.