Using the methods of subtraction clustering and auto-adapted nerve fuzzy network

a sort of motion planner for the underwater vehicle is designed.According to the character of the input and the output

using the methods of subtraction clustering

fuzzy system’s initial structure and parameters are made certain of the blindness and randomness of the subject function definition in the fuzzy logic system design and the automatic pick-up of the fuzzy rules are avoided.Nerve fuzzy system parameter resolution for the nonlinear premise parameter and linear conclusion parameter to recognize separately are put forward.The grads descends arithmetic and least square method are adopted respectively to optimize the front and back parameter of the auto-adapted fuzzy reason system.The simulation results indicate that the auto-adapted nerve fuzzy movement plan is better than the fuzzy control in the same simulation environment.