Abstract
The accuracy and fault tolerance of filters are directly affected by the filter architecture and algorithm, thus influencing navigation performance. The chi square detection used in the conventional reset federated filter is not sensitive to soft faults, and it is easy to cause the health subsystem to be polluted through information sharing. It is a challenge to design an adaptive reset federated filter to improve the performance of the navigation system. Therefore, taking the Strapdown Inertial Navigation System/Global Positioning System/Celestial Navigation System/Synthetic Aperture Radar (SINS/GPS/CNS/SAR) integrated navigation system as an example, an adaptive federated filter architecture for vector-formed information sharing without a fault isolation module is designed in this paper. The proposed method uses the two-state chi square detection algorithm to calculate the parameters corresponding to each state, making the state with higher accuracy obtain a greater information distribution coefficient. In addition, according to the value of vector-formed information sharing, an adaptive coefficient of measurement noise is designed. This improves the adaptability of the navigation system to soft faults. Simulation results show that the accuracy of the proposed algorithm has the same performance compared with the conventional method under normal circumstances. When the sensor has a soft fault, the adaptive federated filter algorithm proposed in this paper can adaptively adjust the distribution coefficients, eliminate the influence of the fault information and improve the precision of the navigation system. The approach described in this paper can be used in multi-sensor integrated navigation. It will have better performance in engineering applications.
Publisher
Cambridge University Press (CUP)
Subject
Ocean Engineering,Oceanography
Reference22 articles.
1. A new method of information sharing in federated filter;Liu;Journal of Chinese Inertial Technology,2001
2. Federated square root filter for decentralized parallel processors
3. Robust Kalman filtering based on Mahalanobis distance as outlier judging criterion
4. Fokin L. , Shchipitsyn A. and Shtessel Y. (2015). Adaptive SINS/ANS/GNSS for Air-Launch Space Launcher: Algorithm Design and Performance Analysis. AIAA Guidance, Navigation and Control Conference and Exhibit, 6760–6779.
Cited by
14 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献