Application of GA to design LQR controller for an Inverted Pendulum System

This paper apply genetic algorithm (GA) to design weighting matrices in linear quadratic regulator (LQR) for an inverted pendulum system (IPS). Feedback gain settings of the system are obtained by minimizing the performance index using GA to optimize the weight matrices of LQR. The divisional of the searching space mechanism and the dynamic crossover and mutation rate are used to insure getting the global optimal results. Time specification performances are also compared considering LQR-GA and experiential-LQR [6] by determine the control strategy delivers with respect to pendulum´s angle and cart´s position. The simulation results show that the proposed method gives the response better than experiential-LQR [6]. All computations will be carried out using MATLABreg.