关键词: 径向基函数, 模糊神经网络, 水下机器人, 状态监测系统

Abstract: A condition monitoring system based on the RBF networks and FNN for the sensors and thrusters is proposed for guaranteeing the safety in task and improving the intelligence degree of underwater vehicles. According to the analysis of the pre-treatment signals, the sensors monitoring model detects the fault of sensors, and restores the signals of fault ones, which are provided as the actual running condition parameters of AUV for the thruster monitoring model. For the corporate action of the actual outputs of the thruster and sensors, the condition information of the thrusters is gained by the evaluation model, and it also realizes the faults location. The results of the computer simulated by actual experiment data of a certain AUV prove that the condition monitoring system proposed is effective and feasible.

Key words: Condition monitoring system RBF, FNN, Underwater vehicle