Determination of LQR controller parameters for flexible link manipulator using meta intuitional algorithms

This study presents analyzes on both the LQR controller designed for motion control of the flexible robot arm system and the optimization of the control parameters. With this optimization work, the flexible robot arm system reaches the desired angular target in the fastest way, while residual vibrations are eliminated. The vibrating particle system is a new meta-heuristic approach and was used for the first time in this study to adjust the parameters (weight matrices) of the LQR. The effectiveness of the method to be used in the study was compared with algorithms such as genetic and artificial bee colony algorithm, which are among the most frequently used optimization algorithms. In addition, this study proposes a new multi-objective function that takes into account the parameters of the control response for the system used. The parameters of the optimization algorithms were determined by means of a large search space. Each algorithm was examined with four different population/particle values and results were obtained for 100 iterations. The best control results obtained from GA, ABC and VPS algorithms are applied to the flexible robot arm system. After the application, the theoretical and experimental results were compared. As a result of the comparison, it was seen that the best control results were obtained with the VPS-based LQR controller. In addition, this study is organized to show the details of the theoretical application in optimization algorithms.