The Research Of Key Technologies Based On Multi-sensor Information Fusion

Multi-sensor information fusion technology has a wide application in multi-mode guidance system, one of the main applications is muti-sensor fusion technology which integrates measurements from multiple sensors information and processes the information through data preprocessing, data correlation, data decision-making, and finally multiple sensor information fusion. The application of multi-mode not only enhances the precision and accuracy but also has a ideal result in practice.The part of data preprocessing, introduced time alignment algorithm for least squares, used effective outliers elimination algorithm and Kalman filtering algorithm, at last,put forward a gray outliers associated methods. Some related simulation was carried out at the same time.The part of data association, discussed the probabilistic data association algorithm, the joint probabilistic data association algorithm. On this basis, the problem of track correlation and in the related is researched, the Hungarian algorithm is used to theoretical simulation, it makes the correlate results more accurate.The part of data decision-making, introduced some commonly used method of D-S evidence combination and compared the advantages and disadvantages of them, on this basis, a new data decision-making method is raised, the comparison with other methods is made to verify the new method that can make a good result.The part of data fusion, introduced the sequential hierarchical fusion algorithm and the optimal fusion algorithm, each level of the two fusion methods are compared, it is obvious to see, the optimal fusion algorithm is superior to sequential hierarchical fusion method.In practical application, the hardware of information fusion is debugged and tested, the interface of the host computer was prepared during the time, it can deal with the measurement data of each part and give the output curve of the corresponding treatment. From the comparison, we can see the simulation algorithm of this paper improved the accuracy and credibility of multi-mode composite guidance.