Design And Analysis Of Filtering For SINS/GPS Integrated Navigation System

With the development of the national defense technology and the national economy, themajor requirements for navigation system is high accuracy, high reliability, low cost andcontinuous working ability for a long time. So far by now, there is no single navigationtechnology can fulfill the demand. In order to improve the accuracy and reliability, there willbe a inexorable trend in navigation area that we combine SINS with GPS or other navigationsystem to form a integrated navigation system. The integrated navigation system deal with theinformation from all sorts of subsystems comprehensively and make full use of theiradvantages to promote the overall performance of the system. In the integrated navigation, theperformance is effected by various errors directly. High precision and the anti-interferenceability can achieve with good data processing method. So it’s distinctly necessary to researchthe filtering method for error processing on integrated navigation system. In this paper, weresearch many filtering method in the integrated navigation system, including Kalman filter、federated Kalman filter and adaptive filter. The main research work of this paper is as follows:Firstly, the paper analyzed the basic principle of SINS and GPS, and established theirerror models. In order to verify the validity of the strapdown algorithm, a implementation ofthe error simulation for SINS was finished through a computer simulation program.Secondly, various combinations based on SINS and GPS were introduced in this paper,and a combined error model of position and velocity in SINS/GPS was designed aftercombing with the principle of Kalman filter. A computer simulation explains that the GPS cansuppresse the growth of navigation error with time and improve the navigation accuracy, andit also verifies the effectiveness of the algorithm in the integrated navigation which based onSINS and GPS.Thirdly, aiming at the large amount of calculation and poor fault tolerance of thecentralized Kalman filter, this paper research a federated Kalman filter algorithm anddesigned a federated Kalman filter without feedback. It was used in the integrated navigationsystem and verified by the Matlab simulation.Last, in the practical application, system noise and measurement noise of the integratednavigation system are unknown, so we designed a federated and adaptive Kalman filter with aadaptive algorithm in the integrated navigation. In order to get its result in the the integratednavigation system, we verified it with the Matlab simulation. The result indicate that thealgorithm is right.