Research On Reliability Of AUV Intelligent Decision System

The reliability of Autonomous Underwater Vehicle (AUV) is one of the important technical specification in measuring the AUV system. This paper studies systematically on how to enhance the reliability of AUV system. The following aspects are investigated and discussed.Firstly, as for the hardware structure and software structure of AUV and uncertainty and complex association in the fault diagnosis of AUV system, a method based on Bayesian Networks is proposed, which is applied into fault diagnosis of AUV. By the inference mechanism based on Bayesian Networks, a modeling process of fault diagnosis system of AUV and the optimizing method of fault diagnosis strategy are given.Secondly, to the architectures and characteristics of subsystems of AUV, this paper presents the reliability design of AUV intelligent decision-making system as the purpose of eliminating the effect by the faults, which is described in generation of fault detection technology , fault tolerant technology and self-recovery technology. The real achievement is introduced in detail.Finally, a method of reliability evaluation based on software and hardware multiplexed system is adopted. On the basis of calculating the reliability of intelligent decision-making system with single computer, the reliability of intelligent decision-making system based on Hot Standby in the process of a mission is obtained according to the reliability analysis of the whole intelligent decision-making system.Results of testing prove that these methods highly improve the ability of fault diagnosis of AUV and the strain ability and reliability of AUV intelligent decision system. The system fits the real design requirement of AUV.