Study On Multisensor Track Association And Fusion Algorithm

Multisensory i nformation fusion technology can combine elements of raw measurements from different sensors into a single set of meaningful information to extract the greater amount of possible information than is present in any individual sensor, and supply the accurate information of target state for optimum battle effect.Two new correlative algorithms are proposed for networking system in this paper, which are track association and estimation fusion based on 2D radar/infrared and 2D Radars, to solve the problems of the weak ability to anti-jamming, the incompletion of measurements and the large computation cost, and to enhance the correct association rate and meet the real-time performance of multisensory information fusion system in battlefield. Both of them adopt two layers correlative algorithms: primary correlation and refined correlation based on Nearest Neighbor(NN). Utilizing the radar contacts, tracks and historical information, primary correlation is operated firstly to reduce the number of correlative tentative while acquire estimation of target state, and then, correct correlation is determined after refined correlation and accurate estimation of target state can be acquired by track fusion. These methods can overcome the difficulty resulted from the incompletion of measurements in fusion processing and enhance the correct association rate efficiently. In addition, in radar/infrared system, radar is controlled by the quality control component, reducing the number of boot-strap of radar while meeting the requirement of tracking accuracy, to ensure covered and safe capability.