Neuron-Fuzzy Planner for a Parallel Parking Maneuver of a Differential Robot

In the automotive area, mobile robot theory with artificial intelligence had demonstrated an incredible potential and interest of several researchers that are involved on the issue of board intelligent systems on passenger cars to perform daily jobs as, navigation without collision, control vehicle speed, parking, etc. This paper will present a path planner using a neural controller trained based on potential fields, in order to determine the trajectory and a fuzzy controller to determine the differential robot kinematics for the parallel parking maneuver task. A simulator presents the neural network training and the fuzzy logic application, from any initial point. The trajectory results can be compared with the theory and the vehicle kinematics can be verified through the simulator.