Study Of Information Fusion Technique In UAV Integrated Navigation Using Payloads

In this paper, which is set in the study of the navigation system of Long Endurance Uninhabited Aerial Vehicle(LEUAV), information fusion theory of UAV-payload-integrated navigation system is researched, for the purpose of approaching the key technique of involving UAV imaging payloads in the integrated navigation system.Multi-sensor information fusion is the leading direction in UAV navigation system development. The information fusion theory of the navigation system is studied in depth in this paper. The three typical fusion techniques, centralized Kalman filtering, decentralized Kalman filtering and federal Kalman filtering, are analyzed, and a new fusion technique called "Error-based Tandem Kalman Filter" is raised.Two information fusion structures are designed for UAV SINS/GPS/payloads integrated navigation system. One is federal Kalman filtering structure, while another is tandem filtering structure raised for the payloads to aid the existing SINS/GPS integrated system. The two filters are designed, and the filtering models of which are constructed. The measurement model for the payloads to aid the navigation system is built. The method of georeferencing using UAV payloads is presented. The application of "two-stage estimation idea" with the ground object's position unknown into the integrated navigation is studied.The information fusion models and methods of the UAV SINS/GPS/payloads integrated navigation system are simulated. Conclusions are drawn via analyzing the simulation results, described as follows: UAV-payloads will improve the navigation accuracy considerably, especially for the yaw angle;During GPS dropout, the system will still keep a good precision with the payloads aiding INS;Error-based Tandem Kalman Filter is easy to realize in structure, which will meet the needs in precision well;Using "two-stage estimation idea", payloads will also do help to the navigation system.