Design and construction of INS/GPS navigation system based on adaptive Kalman filter algorithm

Due to the importance of navigation accuracy, the inertial navigation system is often combined with one of the other navigation systems. In one of these methods, the Inertial Navigation System is combined with Global Positioning System. An integrated navigation system consisting of both systems provides reliable, accurate, and consistent navigation capabilities. The design of the estimator filter is one of the basic steps in the implementation of integrated navigation systems. Due to the difficulty in obtaining the accurate model of nonlinear systems, also the complexity of noise in practical environments and existence of noise with uncertain statistical characteristics, the accuracy of the KF estimation is greatly reduced. Therefore, in this paper, to improve the integrated navigation performance, a concept of self-adaptation to the KF is introduced, and the modified Sage-Husa adaptive Kalman filter algorithm based on the recursive noise estimator basis of the maximum posterior likelihood estimation is formed to overcome the shortcomings of the KF methods and solve the problem of state estimation in practical environments with complex noise and uncertain statistical characteristics and uncertainty in the model. A vehicle test was used to evaluate the proposed algorithm, and the results showed that the proposed algorithm has very acceptable accuracy and performance. so it was able to improve the RMSE evaluation criteria of position in the direction of height and speed in the vertical direction by about 38% and 25%, respectively and The RMSE evaluation criteria and Std evaluation criteria of the heading angle are 18% and 17%, respectively.