Research on AUV Obstacle Avoidance Based on BP Neural Network

摘要

Autonomous underwater vehicle should have real-time obstacle avoidance ability of self-protection in autonomous operation in unknown environment. A three-dimensional real-time obstacle avoidance method is proposed for under-actuated AUV with the distance sensor as obstacle avoidance sensor. The output information of the distance sensor is converted into a dangerous degree which as the input of BP neural network. The output of BP neural network is the heading or the depth of AUV which is to be adjusted. The effectiveness of obstacle avoidance method based on BP neural network is verified by MATLAB simulation.