Research On Fault Diagnosis And Tolerance Control For AUV System

At the recent years, with the development of ocean research and exploitation and the need of national defence, higher request is put forward for the intelligent level of Autonomous Underwater Vehicle(AUV), which is an important component of ocean high-tech. As the condition which AUV works in is bad and the environment is complex and changeful, it is important to design a reliable control system to ensure. it work safety. It is an important embodiment of intelligence to realize automatic fault diagnosis of AUV control system.The purpose of this thesis is to realize automatic fault diagnosis and fault tolerance of AUV control system, to prevent control performance degraded and fault extended, and then to improve the viability and motion performance of AUV at complex sea circumstance.In this thesis, the sensor fault tolerance algorithm for sensor faults and actuator fault detection algorithm for actuator are discussed. For the sensor fault, in the basis of analyzing a decoupling controller and sensor system of an Autonomous Underwater Vehicle (AUV), fault-tolerant control strategy based on the ideas of using outputs of sliding-mode observer to substitute outputs of fault sensor is proposed. To the irrecoverable sensor fault, this strategy minimizes its damage to AUV. To the jumping fault of sensors, substituting control is used in the time of jumping. The simulation result shows the validity of this strategy. For the actuator fault, the neural networks algorithm is focused. Based on the character that hidden layer wavelet function of wavelet networks can adjust scale factor and shift factor to affect output of networks, the least disturbance algorithm adding scale factor and shift factor is proposed and it can calculate the dynamical learning ratio to improve the stability and convergence of wavelet networks. This algorithm is applied to build the dynamical model of autonomous underwater vehicle (AUV) and via comparing the output of neural network with the real state value in the condition of thrust fault, fault detection rules are distilled to execute thrust fault diagnosis. The result of simulation proves the effectiveness of this algorithm.As the experiment results show, the sensor fault tolerance algorithm and actuator fault detection algorithm presented in this thesis can detect and tolerate the fault of AUV control system on-line, which can provide precondition for the building of farther fault recovery and tolerant control system, and is very useful and significant in the field of AUV.