The Research On Information Fusion Of Compound Guidance

With the constant development of the military technology, multi-mode compoundguidance technology has been become the main developing direction of homing guidanceweapons. Information fusion technology is one of the key technologies for compoundguidance system. The main function of multi-sensor information fusion technology is used toanalyses the measurements from multiple sensors. Through the data preprocessing, datacorrelation, data decision-making and date fusion for the multiple sensor information,multi-sensor information fusion would be produced a more accurate and reliable result thansingle sensor does. The correlative theory and application of multi-sensor information fusionin compound guidance were researched in this paper.In the part of data preprocessing, the main algorithm of time registration and spaceregistration was introduced in this paper. And about outlier elimination, Kalman filteringalgorithm, least squares algorithm and date fluent movement algorithm were researched. Thenthe related simulation was carried out in the meantime.In the part of data association, some most basic and simple methods like nearestneighbor data association, K-NN arithmetic were introduced. The probabilistic dataassociation algorithm and joint probabilistic data association algorithm were researched as thekey problems, and simulation results comparison was presented.In the part of data decision-making and data fusion, this paper introduces the traditional D-Scombination evidence theory, but it would give a unreasonable conclusion when the highlyconflicting evidences exist. Therefore, this paper researches some modified methods also, anda new algorithm of evidence combining was introduced, the simulation results contrast showsthat they have better performance on data decision-making. About data fusion, hierarchicalfusion was mainly researched, and the simulation of sequential hierarchical fusion algorithmis carried out.