| Unmanned Underwater Vehicle(UUV)plays an important part in the development and exploration of marine resources and is an important part of underwater intelligent equipment.Among them,the stability and safety of UUV mission navigation is the key and foundation of underwater work.UUV sensors and actuators are the most important operating parts of UUV.The normal operation of sensors and actuators is the burden of smooth completion of UUV mission navigation.The research on fault diagnosis and fault tolerance control of UUV sensor faults and actuator faults,solving the track deviation caused by UUV faults,and completing the self-rectification of the track during the UUV mission are of great value to the UUV’s underwater navigation work.In this paper,the UUV track deviation problem caused by DVL distortion fault and thruster fault is studied from three aspects: DVL fault detection and reconstruction,thruster thrust loss fault diagnosis and fault degree identification,and UUV track self-rectification under multiple faults.DVL fault detection and reconstruction problems.For the parameter optimization algorithm of the generalized regression neural network,the fruit fly optimization algorithm needs to jump out of the local optimal solution and improve the traversal ability of the solution space.The flock optimization algorithm using the strong and accurate chicken flock algorithm idea is performed.Improve.Aiming at the DVL distortion failure caused by the complex magnetic field inside UUV,a neural network was used to construct the DVL prediction signal,and the DVL signal distortion was detected by calculating the sample entropy of the DVL output signal.The results show that the sample of the signal is calculated when analyzing the complex changing DVL output signal Entropy can accurately distinguish the distortion state of DVL signals,and can predict the speed of UUV under the prediction of neural network.Diagnosis of thrust loss failure of thruster and identification of failure degree.Aiming at the problem of the complex internal magnetic field environment of UUV,which severely interferes with the control voltage and measured current of the thruster and affects the application of the Bayesian algorithm,the wavelet optimal scale reconstruction method is used to reduce the impact of external interference.The optimal scale is calculated and determined by the sample entropy algorithm.Simulation results show that using the wavelet optimal scale reconstruction method to analyze the current signal of the UUV thruster thrust loss fault can accurately and efficiently detect the fault characteristics of the fault;the wavelet optimal scale reconstruction of the voltage signal can obtain thrust loss fault characteristics under Bayesian algorithm clearly.UUV track self-rectification under multiple faults.Aiming at the impact of external current interference on the DVL sensor and the deviation of UUV under emergency navigation.the current information during UUV navigation was estimated by the current estimation principle,and the current interference of DVL reconstruction signal was corrected by the DVL fault detection and signal recovery algorithm,which realizing UUV emergency navigation and self-rectification under DVL failure.Aiming at the UUV navigation trajectory deviation caused by the thrust loss failure of the thruster,the UUV heading angle and longitudinal speed closed-loop control were designed respectively for the UUV heading angle integral S-plane controller and longitudinal speed integral S-surface controller to cooperate with the thruster failure Diagnosis and identification algorithm,fault-tolerant control of thrust loss failure of UUV thruster,realizing UUV direct self-rectification navigation. |
