Fuzzy logic in auto-tuning of fractional PID and backstepping tracking control of a differential mobile robot

The use of the fractional PID controller or simply (PID mu)-D-lambda has brought the addition of two new parameters, lambda and mu. Although this results in a greater flexibility, they make the tuning of the controller more complex and slower. One solution to this is the use of fuzzy logic to perform a self-tuning of the parameters. Through its rules of inference, the algorithm can determine a better tuning in real time. This article presents a practical application of a PI lambda self-tuned with the use of fuzzy logic, in a differential mobile robot. Three different types of speed controllers are presented. A lemniscate curve is used for the robot's trajectory tracking, with and without disturbance in the speed control of the wheels. Dynamic selection of better controller parameters is obtained by the fuzzy controller and applied on the speed control of the wheels of a mobile robot. A camera is used as feedback for the tracking controller, so the real pose estimated is acquired through image processing. The self-tuning controllers are evaluated, compared to a fixed-tuning controller, with its parameters being acquired through traditional, well consolidated methods. The implementation, practical results and conclusions are hereby presented.