Tracking Control for AUV by Fuzzy Controller with Extended Kalman Filter

Abstract

In this paper, we verify whether a fuzzy control system designed for a small autonomous under water vehicle (AUV) robot can be controlled by inputting state estimates obtained by using an Extended Kalman filter. The fuzzy controller determines a set of rules similar to those used by humans to steer the robot, and quantifies them in the form of a membership function to obtain a control value. The state estimator uses an EKF to reduce noise from noisy velocity measurements. The current position, which cannot be measured, is estimated and input to the controller for control simulation. As a result, the AUV succeeded in tracking the target value moving in three dimensions in time.