Fault Tolerant Algorithm Research And Simulation Realization For Multi-sensor Integrated Navigation Systems

With the rapid development of electronic information technology, the modern battlefield environment of high altitude was growing more complex. A variety of radio navigation systems including GPS, Compass, GLONASS and Galileo navigation system, were vulnerable to enemy electromagnetic interference; Celestial navigation system was vulnerable to the impact of natural environment; Terrain aided navigation was limited by the region and the capacity of computer storage; These adverse factors lead that the navigation sensors didn't work well, and the output of navigation information was wrong, which affected the accuracy and reliability of navigation systems. According to the application background and work characteristics of the high altitude and long endurance UAV(unmanned aerial vehicle), its navigation system should have high reliability and high navigation precision for long time, so a single navigation model could not meet the needs for its work.. To this end, this paper designed the corresponding integration program of SINS/multi-sensor integrated navigation system, and studied the integration algorithm and fault detection algorithm in depth. On this basis, the simulation platform of SINS/multi-sensor integrated navigation system was designed, so as to set some theoretical basis for project application.Firstly, considering the application background and work characteristics of UAV, the paper studied the fusion algorithm of SINS/GPS/CNS/SAR integrated navigation system in detail; According to the error characteristics of navigation sensors, the paper established the corresponding state equation, measurement equation, information distribution coefficient, and then created the corresponding kalman filter model; The result of MATLAB simulation proved the fusion algorithm was feasibility and effectiveness.Secondly, the paper studied the sensor fault detection algorithm. Considering the disadvantage of residual chi-square detection algorithm which had low sensitivity to small fault information, the paper proposed a improved fault detection algorithm based on eigenvalue detection; The result of MATLAB simulation showed that the algorithm was better than the residual chi-square detection algorithm, and the improved method could further improve the accuracy and reliability of navigation systems.Finally, the paper designed a simulation experiment platform of SINS/multi-sensor integrated navigation system based on the distributed network. The experiment platform consisted of inertial sensor simulation subsystem, multi-sensor simulation subsystem, fusion and control subsystem, and display subsystem. The experiment platform could research a variety of integrated navigation algorithm, and had strong flexibility and adaptability.