Research On Integrated Navigation System Of High Altitude / Long Endurance Uavs Base On The Astronomy Sensor

High-altitude/long-endurance(HALE) UAVs put forward higher demands for the autonomy and accuracy of long-endurance navigation which single navigation system can not achieved. Integrating of INS and CNS could give full play to their strengths and compose high-precision navigation system which has strong autonomy and anti-interference ability. To propose celestial navigation attitude determination, positioning and INS/CNS integrated navigation algorithm, integrated navigation technology based on astronomical observation information for HALE UAVs was studied.For the characteristics of HALE UAVs, celestial navigation attitude determination and positioning algorithms and their error characteristics were studied deeply. For the conventional TRAID algorithm has asymmetric for the two vectors, an improved TRAID algorithm which effectively improved the accuracy of attitude algorithm was presented. Based on height difference location algorithm and improved TRAID algorithm, impaction of their error characteristics to CNS was researched detailly. These works provide theoretical support for following simulation study.Combining the stable flight characteristics of HALE UAVs, the INS/CNS filtering algorithm was studied. For height has non-observation in astronomical position, height aided INS/CNS position integrated navigation system was designed. The attitude which star sensor output in the inertial coordinate can not be directly applied to the actual geographic coordinates, so this paper introduced position and attitude error transformation matrix, and established the coupling error model between attitudes in the two coordinates. Corresponding integrated filtering algorithm was proposed.HALE UAVs have the characteristic of multi-sources of information and Beidou system would be applied in the future. So satellite navigation was introduced to construct the INS/CNS/GNSS integrated navigation system based on federated filter. Fading sage adaptive filter algorithm was proposed. The measurement noise was online adaptive estimation in this algorithm. Fading factor was introduced to restrict the length of filter memory and enhance the tracking ability of filter.Based on VC++, this paper designed and completed a distributed simulation platform of INS/CNS integrated navigation. The platform has good scalability and maintenance. Various modes of INS/CNS integrated navigation algorithm can be simulated and analyzed through the platform. So it can provide a good platform for INS/CNS integrated navigation algorithm validation.