The aim of this paper is to compare the performances of a fractional order proportional integral derivative (FOPID) controller tuned with evolutionary algorithms for robot trajectory control. In order to make this comparison, a 2-degrees-of-freedom planar robot was controlled by a FOPID controller tuned with particle swarm optimization (PSO) and a real coded genetic algorithm (GA). In order to see the effects of the cost functions on the optimum parameters of the FOPID controller, 3 different cost functions were used: the root mean squared error (MRSE), mean absolute error (MAE), and mean minimum fuel and absolute error (MMFAE). In order to compare the performance of PSO and the GA under different conditions and to test the robustness of the FOPID controller tuned with these algorithms, the parameters of the system model and the given trajectory were changed and white noise was added to the system. All of the simulation results for the robot trajectory experiment show that the FOPID controller tuned by PSO has better performance than the FOPID controller tuned by the GA. Furthermore, the results obtained for the FOPID tuned by both PSO and the GA indicate the superiority of the proposed tuning approach for robot trajectory control.

The aim of this paper is to compare the performances of a fractional order proportional integral derivative (FOPID) controller tuned with evolutionary algorithms for robot trajectory control. In order to make this comparison, a 2-degrees-of-freedom planar robot was controlled by a FOPID controller tuned with particle swarm optimization (PSO) and a real coded genetic algorithm (GA). In order to see the effects of the cost functions on the optimum parameters of the FOPID controller, 3 different cost functions were used: the root mean squared error (MRSE), mean absolute error (MAE), and mean minimum fuel and absolute error (MMFAE). In order to compare the performance of PSO and the GA under different conditions and to test the robustness of the FOPID controller tuned with these algorithms, the parameters of the system model and the given trajectory were changed and white noise was added to the system. All of the simulation results for the robot trajectory experiment show that the FOPID controller tuned by PSO has better performance than the FOPID controller tuned by the GA. Furthermore, the results obtained for the FOPID tuned by both PSO and the GA indicate the superiority of the proposed tuning approach for robot trajectory control.