Fuzzy Logic Path-Tracking of Car-like Mobile Robots based on Curvature Derived Point Selection Algorithm

In the path tracking problem of the car-like mobile robots, selection of tracking point and efficient control system are the fundamental steps of the overall control system design procedure. In this paper curvature-derived point selection (CPS) based on kinematic constraints of the car-like mobile robots is presented in order to select an appropriate tracking point. According to the presented algorithm, the selected point is the location of the path where the possibility of the tangential motion relative to the path exists for the robot. By this algorithm, according to analyses of the forthcoming path of the robot, the appropriate steering reaction is expected from the robot where the slope of the path varies discontinuously. Afterwards, the selected point is given to an appropriately designed fuzzy logic control system in order to be tracked accurately. The presented tracking algorithm is numerically simulated and is compared to similar fuzzy-based method. The obtained results indicate that for tracking a desired straight line, the normalized steering activity factor (NSAF) is reduced from 0.1259% to 0.0030% for similar fuzzy-based method and proposed algorithm, respectively. Also the normalized average relative error (NARE) of the robot for tracking a polygonal path is reduced from 29.6762% to 16.6314% for similar fuzzy-based method and proposed algorithm, respectively.