基于BP神经网络的水下机器人推进器故障辨识算法  

A Fault Identification Algorithm of Thrusters for Unmanned Underwater Vehicles Based on BP Neural Networksn

作　　者：陈艾琴[1] 刘乾[1] 朱大奇[1] CHEN Ai-qin, LIU Qian, ZHU Da-qi （Laboratory of Underwater Vehicles and Intelligent Systems, Shanghai Maritime University, Shanghai 200135, China）

机构地区：[1]上海海事大学水下机器人与智能系统实验室,上海200135

出　　处：《电脑知识与技术》2009年第7期5214-5216,共3页Computer Knowledge and Technology

基　　金：上海市教委课题（NO.2008099,NO.20080119）

摘　　要：常规无人水下机器人推进器故障诊断中，均假设推进器处于几种固定故障模式，这与实际推进器故障情况有较大差别。该文将信息融合故障诊断技术引入推进器拥堵故障在线辨识之中，提出基于BP误差反传神经网络（Error Back Propagation Network）信息融合在线故障辨识模型，将水下机器人控制信号和故障情形下的方向偏转率作为BP神经网络融合模型输入，其输出即为反应推进器故障大小的拥堵系数，不仅提高了故障辨识精度，而且对连续不确定故障实现有效辨识。Normal state and several different fault patterns are considered in conventional thrusters fault diagnosis of unmanned underwater vehicles, the control law is designed off-line. But it is not different from actual thrusters fault situation. In this paper,information fusion fault diagnosis technology, has been the introduction of congestion thruster fault line identification, and information fusion line fault identification model based on Error Back Propagation Network has been raised. Underwater robot control signal and the direction of deflection rate in fault situations as input of BP fusion model, and the Congestion factor reaction thruster fault size is its output. This approach not only improves the accuracy of fault identification, and achieves effective identification to continuous uncertainty on failure.

关 键 词：水下机器人 推进器 BP神经网络 故障辨识 信息融合 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论;自动化与计算机技术—计算机科学与技术]