Research On Polar Federated Filter Algorithm For Low Quality Observation

High precision and high reliability polar integrated navigation is a prerequisite for all kinds of military and civilian marine carriers to achieve polar arrival,polar navigation and polar operation.However,the periodic oscillation error of grid inertial navigation system,the position error accumulated over time and the low-quality output of auxiliary navigation system affected by the environment will have a great impact on the performance of polar integrated navigation.Low quality observation refers to the situation that the output information of the auxiliary navigation system contains fault signals or the output information is missing for a short time due to the influence of the polar environment.How to improve the integrated navigation performance under the condition of low-quality observation in the polar region is an important problem to realize high-precision and high reliability polar navigation.Therefore,this paper studies the polar federated filtering algorithm under the condition of low-quality observation,in order to solve the problems and technical difficulties of the current polar integrated navigation,and improve the accuracy and reliability of the polar navigation system.The main work of this paper is as follows:Analyze the limitations of North pointing inertial navigation in the polar region,study the mechanical arrangement scheme and error equation of grid inertial navigation,verify the feasibility of grid inertial navigation in the polar region,periodic oscillation error and position error accumulated over time through simulation experiments,analyze the observability of grid inertial navigation error model,and verify the analysis conclusion through simulation experiments,According to the analysis conclusion,the measurement information of the sub filter is selected to provide theoretical basis and support for the subsequent design of reset federated filter.Aiming at the problems that the analysis of polar satellite navigation performance is not comprehensive and the analysis conclusion is not representative in the existing literature,based on the precise ephemeris,this paper makes a comprehensive analysis on the visibility and positioning accuracy of the current constellation of the global four David navigation system in the polar region through the methods of route point selection and regional analysis;Aiming at the problems of low accuracy and poor reliability of auxiliary navigation system,a SINS/DVL/GNSS/CNS integrated navigation system framework based on reset federated filtering algorithm is proposed.The simulation experiment verifies the improvement effect of the framework on the accuracy of navigation parameters.Aiming at the problem that the output information of auxiliary navigation system in federated filter may contain fault signal,the detection effect of residual chi-square test on fault signal is studied,and a residual chi-square test and fault-tolerant algorithm based on fuzzy logic is proposed to improve the problem of large delay of residual chi-square test soft fault,so that the fault detection algorithm can be sensitive to soft fault earlier,The information distribution coefficient of the federated filter is adaptively adjusted and the federated filter is reconstructed to reduce the influence of soft fault on the federated filter.The fault tolerance of federated filter is improved,and the effectiveness of the algorithm under soft fault is verified by simulation experiments.Aiming at the problem that the auxiliary navigation system does not output information in a short time in the harsh environment of the polar region,the influence of this problem on the output error of the federated filter is studied,and a discontinuous measurement information estimation algorithm based on long short term memory is proposed to solve the increase of the federal filter error caused by the lack of measurement information.And the algorithm solve the problem of at the time of lack and recovery of measurement information,the federated filter has the problem of error jump,which finally improves the accuracy and reliability of the integrated navigation system.