An intelligent control for a crawling unmanned vehicle

Describes a method to design an adaptive feedback controller for an unmanned crawling vehicle (UCV) having track style wheels. The motion of the vehicle is constrained by environmental uncertainties, which are represented by unknown friction coefficients between the tracks and ground surface. The controller, a combination of conventional PID and adaptive neural network fuzzy logic (ANNFL), is designed by using multiobjective optimization techniques from which the fixed and the adjustable control gains are obtained. Using the nonlinear motion equations of the vehicle dynamics, the PID controller is designed at a nominal operating point and establishes a desired system response pattern of feedback (closed loop) control.. The ANNFL deals with the uncertainties to compensate the PID controller for reducing the "command following error" by tuning an additional gain feedback from the system output besides the fixed gain of the PID controller. ANNFL provides an adaptability and robustness for the integrated control strategy of the PID and ANNFL. An example simulation is included