Research On Signal Correlation And Information Fusion Of Distributed Sensors

With the development of radar technology system,computer technology and other disciplines,the form of information in the electromagnetic environment is changing rapidly.Therefore,it is very meaningful to use multi-sensors to perform multidimensional detection,correlation and information fusion on the target to obtain complete and accurate information.This paper mainly proposes electronic support measures(ESM)and radar data correlation methods based on evidence theory for heterogeneous multisensor systems.A track correlation scheme based on data correlation for passive multisensor systems is proposed and verified by simulation analysis.Based on the theory of information fusion,this article focuses on the following four aspects.1.For heterogeneous sensor fusion systems,this article mainly studies the use of bearing information to correlate ESM data with radar tracks.This paper proposes a improved method fusion rules,gives a function that generates basic probability assignments,and proposes an association method based on D-S evidence theory.The correlation performance of the algorithm with the maximum likelihood-based correlation method and the fuzzy synthesis-based correlation method is compared through simulation,and the universality of the algorithm for different ESM and radar sampling cycle ratios,and the superior correlation performance for azimuth close target.2.In this paper,the fearture data correlation problem of fixed fearture radar is studied.The correlation function in Gaussian model is given.The basic sequential algorithm scheme(BSAS)is improved and applied to the radar attribute correlation problem.Aiming at the problem that BSAS is sensitive to the order of data participation,BSAS algorithm combined with simulated annealing algorithm and BSAS algorithm combined with ant colony optimization algorithm were proposed.By comparison with BSAS and BSAS based on redistribution,it is proved that the proposed algorithm can improve the order sensitivity of BSAS and improve the correlation performance.3.This paper studies the fearture data association problem of time-varying fearture radar.An association function based on set theory,and mathematical expressions and approximate solutions of association thresholds are given.In this paper,the correlation function is applied to the Density-Based Spatial Clustering of Applications with Noise(DBSCAN)algorithm,and the specific schedule of radar attribute correlation with time-varying attribute characteristics is given.Simulation experiments prove the superiority of the correlation performance of the algorithm under different detection probabilities.4.This paper studies the single target tracking problem under the multi-sensor model.In this paper,the Modified Gain Extended Kalman Filter(MGEKF)algorithm in the single-sensor model is extended to the multi-sensor model,and the MGEKF correctable equation form under the multi-sensor model is derived.The superiority of the positioning performance of the algorithm is proved by simulation.