Study On SINS/CNS/GNSS Integrated Navigation Methods

Navigation technology with high accuracy, high reliability and real-time performance is one of the key technologies in accurate strike and battlefield object detection. So far by now, there is no single navigation technology can fulfill the newest weapon's requirements, but by adopting the SINS/CNS/GNSS integrated navigation method (i.e. navigation method which are combined with Strapdown Inertial Navigation System (SINS), Celestial Navigation System (CNS) and Global Navigation Satellite System (GNSS) navigation method) can fulfill these requirements. SINS/CNS/GNSS integrated navigation can take advantages of all of these three different navigation methods and achieve a much better navigation performance than each of them. The performance of SINS/CNS/GNSS integrated navigation method in theory is pretty good, but in fact, not only it is such a complex task to construct a integrated system, but also the system is sensitive to environmental disturbance, and the information redundancy and information amalgamation are hard to actualize, so that the performance of the navigation system is less than expected. In this thesis, we will concentrate on these obstacles and take medium/long-range ballistic missile and Long Endurance Unmanned Aerial Vehicle (LEUAV) as research objects to analyze the integrated navigation system. By researching on the navigation accuracy, reliability and immediacy performance, a semi-physical simulation platform is built. The main innovation points of this paper are as follows:1.Since the SINS/CNS navigation system could not directly estimate the velocity and position errors of ballistic missile, an error correction method based on status transfer matrix is developed ,which can estimate the velocity and position errors when the missile is at its boost phase. The method mainly use the attitude information provided by star sensor to estimate the initial misalignment angle, gyro drifts and accelerometer fixed bias of the SINS, and combine the status transfer matrix at every occasion of the boost phase, and then with all these information we can invert and correct the velocity and position errors. The simulation results show that this method effectively improve the navigation accuracy of the SINS/CNS integrated navigation system on ballistic missile.2.When the ballistic missile is at its boost phase, its high dynamic specialty demands quite high immediacy performance of the navigation system, and the high dynamic specialty also leads to the nonlinear problem of the navigation model. We develop a SINS/CNS navigation method which is based on dimension reduction model and Ensemble Neural Network arithmetic. By analyzing the observable degree of the states of SINS/CNS integrated navigation system, we can build a dimension reduction model of the SINS/CNS integrated navigation system, and with the help of Ensemble Neural Network arithmetic, which have a strong ability in building nonlinear model, we can amend the errors in building model of navigation system. The simulation results show that this approach has a slightly higher accuracy than UKF method, but it can tremendously reduce the amount of calculation than the full dimension model.3.For the navigation system of LEUAV, its gyro axis misalignment errors and scale factors are varying over time. To solve these problems, a real-time calibration method of SINS/CNS integrated navigation system, which is based on model error prediction, is proposed. The method firstly build up a high dimension model including gyro axis misalignment errors and scale factors errors; secondly, by referring to the observable degree analyzing information, consider some of the states as the model errors and use the error prediction filter to estimate these states; and then, finally, a improvement of the accuracy of the real-time gyro calibration is achieved. The semi-physical simulation results show that by using this method in calibration, the attitude accuracy of SINS/CNS integrated navigation system is improved.4.When SINS/CNS/GNSS integrated navigation system in LEUAV works continually for long time, the effect of disturbance which comes from pneumatic circulation and environmental changes will lead to system model uncertainty and unknown external disturbance characteristic. A SINS/CNS/GNSS integrated navigation method based on H₂/H_∞robust multi-object hybrid filter is developed, which guarantees the celerity of convergence rate and reliability of the system. The characteristic of H₂ filter keeps a low standard deviation of the errors of the navigation system while the characteristic of H_∞filter keeps the errors of the navigation system in a acceptable range, therefore robust multi-object optimization is achieved and the comprehensive performance is improved of the navigation system. The semi-physical simulation results show that the navigation accuracy of this method is better than that of H∞filter.