Research On Filtering And Fault-tolerant Algorithms For A Class Of Integrated Navigation Systems

High altitude long endurance unmanned aerial vehicle(UAV)has a long flight time and a high flight altitude,so it requires its autonomous navigation system with high precision and reliability.Strapdown Inertial Navigation System(SINS)has good autonomy,but its error accumulates with time;Global Positioning System(GPS)has high positioning accuracy,but its signal is susceptible to interference;Celestial navigation system(CNS)has high accuracy of attitude.However,a single integrated navigation mode can not meet the requirements of the HALE UAV.Therefore,SINS/GPS/CNS integrated navigation system is composed by multi-sensor fusion to improve the accuracy reliability and autonomy of the system.Aiming at the high reliability requirement of UAV navigation system,the filtering and fault-tolerant technology of SINS/GPS/CNS integrated navigation system is studied in this paper.The main research work is as follows:Firstly,the basic principles and concepts of integrated navigation system are introduced,and the error model of SINS is established.The SINS/GPS/CNS integrated navigation scheme based on federated filtering structure is constructed.The SINS/GPS/CNS integrated navigation system is modeled,the SINS/GPS and SINS/CNS local filtering subsystem models are established respectively,which lays the theoretical foundation for the following chapters of this paper.Aiming at the inaccuracy of model structure parameters and noise statistical characteristic parameters of integrated navigation system,an improved adaptive strong tracking Kalman filter algorithm is studied.The Sage-Husa adaptive filter and strong tracking filter algorithm are combined to estimate the measurement noise online and real-time,adjust system parameters and prevent filter divergence.UD decomposition is added to the filtering process.The stability of the filter is guaranteed.The simulation results show that the improved adaptive strong tracking filtering algorithm has high filtering accuracy and catastrophe tracking ability.Aiming at the problem of sensor faults during UAV flight,an improved residual chi-square test method based on innovation eigenvalue extraction is studied to detect system faults.In order to reconstruct the system,an adaptive fault-tolerant filtering algorithm based on fault information allocation coefficient is studied.The filter weight is allocated by the fault-free rate of each sensor,and the detection threshold is processed by fuzzy logic.Finally,the output of each sub-filter is input to the main filter according to thepreset information allocation principle,and the global optimal estimation can be obtained after fusion.The simulation results show that the proposed algorithm can detect and deal with faults well and ensure the reliability of the system.In order to further verify the performance of the proposed filtering algorithm and fault-tolerant integrated navigation algorithm,a vehicle-mounted experiment of SINS/GPS/CNS integrated navigation system is carried out,and a simulation acquisition platform of SINS/GPS/CNS integrated navigation system is constructed.The data obtained from the vehicle-mounted experiment are simulated offline to provide simulation verification for the algorithm in this paper.