Federal Filtering Algorithm Of Heading Angle Based On INS/GPS/Magnetometer Integrated System

The integrated navigation system consists of no fewer than two navigation mode. It has much higher performance, more security and stability. There are many integrated navigation system. The different combinations of the satellite navigation, inertial navigation and geomagnetic navigation are the major system. Heading angle is a very important navigation parameter. In this dissertation, the course angle algorithm of INS (Inertial Navigation System)/GPS (Global Positioning System)/magnetometer integrated navigation system has been researched.First of all, this paper has summarized the research actuality of navigation system and the research background and topic basis. It also gave an outline of the INS and geomagnetic navigation system and analyzed the navigation principle. The detailed analysis of their advantages and disadvantages has been expounded. Then the paper introduced kalman filtering algorithm, extended kalman filtering (EKF) and the federal filtering. After that, the hardware of experiment and the experimental process have been introduced in this paper. Some integrated navigation systems related to the research contents have been introduced, too. At the last part, it introduced the federal filtering algorithm which designed in this paper and its results in detail.The last chapter introduced the whole process of the algorithm in detail according to the sequence of research. First of all, some basic researches about magnetic heading have been done and the hard magnetic error of magnetic heading has been removed by using the elliptical effect. Then analysed the error of inertial navigation, such as conic error, paddle error and vortex volume, and compensated these error. Then designed the EKF algorithm for INS/GPS integrated course and the initial precision alignment has been designed by kalman filter. At the end of this part, the centralized filtering and federated filtering of INS/GPS/magnetometer integrated heading have been designed.In the every step of the research, the effectiveness of the algorithm was validated with the measured data and reasons have been analyzed. Relative to the INS/GPS integrated course, its inhibition of error accumulation of inertial navigation is better. The mean square error of heading angle was decreased from1.57°to0.84°. Compared with the centralized filtering algorithm, the fault tolerance of federated filtering is better. The mean square error of heading angle was decreased from0.96°to0.84°. The comparison analysis of these results shows that the federated filtering algorithm has a very good effect:the precision of course is the highest and the stability is the strongest among the several algorithms in this paper.