A Lecture on Current Limiter the Motion Control of Robotic Fish Based on BP Neural Network

Abstract

Robotic fish is a multivariate and nonlinear controlled object, its work environment is complicated, these lead the motion controlling of robot fish be very difficult. This paper provides a method of robotic fish controlling based on BP neural network, and on URWPGSim2D simulation platform. Using the relative position of the fish robot, the ball and the target point in current cycle, this method can compute the two controlling value of robotic fish, which are the velocity and angular velocity, in the next cycle. The experiments show that this method can efficiently reduce the error that comes from the platform randomness, and robotic fish can run smoothly according to the predefined moving route.