Research On Autonomous Navigation With Multi-source Information Fusion Of Long-endurance UAV

With the features of carrying out strategic and tactical tasks for a long time and a large area,Long-endurance UAV is now the research hotspot of the military. The electromagneticenvironment of modern battlefield is complex, and long-endurance UAV has more demand ondependability and stability. The classical INS/GPS integrated navigation System gets affectedeasily by enemy electromagnetic interference, because it needs outer information. Based on theflight characteristics and mission requirements of the long-endurance UAV, the navigation systemindependent of satellite navigation system is researched. Moreover, the simulation experimentplatform is designed for multi-source information fusion, and this provides strong background forfurther engineering application of autonomous navigation.Firstly, based on application background and performance characteristics of thelong-endurance UAV, the intermittent information fusion project of autonomous navigation israised to solve the problem of accessing aided navigation system network. After the research ondata fusion of integrated navigation system, sub filter modification algorithm is designed. Thestate update method of federated filtering is put forward. This method can keep strong faulttolerance and in the same time solve the adjustment problems of traditional no-reset federalalgorithm in applying to in-coordinate interval kalman filtering.Secondly, focusing on the abnormal situation of long-running navigation system, residualfailure detection algorithm of adaptive time delay is raised.. As the inertial navigation system isliable to get divergence along with time, it can cause the failure detection error and isolate thenormal sub-system. So the failure detection is researched in the situation of intermittentinformation fusion to solve related issues. All above ensure the high reliability of navigationsystem on long-endurance UAV.Finally, based on all researches above, the digital simulation system of multi-sourceinformation fusion autonomous navigation is designed to carry out system tests and verification.Using two kinds of system model, comprehensive simulation tests and analysis are made. Thesimulation results show that the autonomous navigation system can meet high precision andreliability. Thus, the long-endurance UAV navigation system algorithm can be implemented forengineering application in the future.