Comparison of performance of pid, fuzzy and gauss controllers in behavior planning of mobile robots

Identifying behavior modeling algorithms to be applied to the differantial drive mobile robot is one of the most important issue in robotic applications. in this context, it is necessary to characterize the robot movements according to the tasks and environment variables in the working environment (configuration space). In this thesis, navigation algorithms which were applied to a differential drive mobile robot trying to reach target by avoiding from the obstacles were examined. Using the data from the sensors on the mobile robot, it is ensured that the robot is able to perceive the surroundings and act according to the environment. The data are obtained from eight ultrasonic distance sensors on the mobile robot. Thanks to the ultrasonic distance sensors, the position of the mobile robot according to the obstacles is determined. These data were used as input value in the algorithms used. With these input values, the avoid behavior of the mobile robot is modeled. In this study, previously developed PID based Sim.I.Am mobile robot simulator has been examined. In this context, Sim.I.Am mobile robot simulator application to eliminate the deficiencies detected and the algorithm used instead of the results were observed using different algorithms. In the first part of the study, the optimization of the PID navigation algorithm applied to the mobile robot is explained. Especially the algorithms used according to the environment where the mobile robot moves can be insufficient. In the following parts of the study, it is proposed to configure the navigation controllers of the differential driven mobile robot systems with Fuzzy and Gaussian algorithms. Sim.I.Am simulation was used to show the effectiveness of the developed algorithm base controller. When the simulation results are examined, it is observed that the algorithms examined and proposed in a simulation test environment where the location of the obstacles change, produce a successful result. Within the scope of the thesis, the performance of the controller obtained by the proposed algorithms is compared with the previous studies.