Passive Localization Algorithm For Acoustic Network Under Dense Clutter

Passive acoustic tracking system is to track low altitude and ultra-low altitudetarget with azimuth information provided by acoustic sensors network. It usesinformation fusion technology to realize the passive target location and tracking.Due to the interference of background noise, the passive acoustic detection networkreceives measurement data contains dense clutter. Clutter processing effectivenesswill affect the final tracking results. Therefore, the research in target location underpassive acoustic network with dense clutter has great significance.Under the General Armaments Department weapons and equipment pre-researchproject, this dissertation researches the data pre-processing and passive acousticnetwork location under dense clutter to realize the target location. The experimentaldata shows the efficiency of location algorithm. The main work is given as follows.First, for the problem of acoustic sensor network low altitude and ultra-lowaltitude tracking in dense clutter, a clutter pro-processing algorithm is proposed. Ituses the nearest neighbor association and the iterative way to eliminate clutters inraw data and extract target information. In addition, when dealing with multipletargets, the algorithm uses the nearest neighbor principle and fitting method toeliminate clutters in raw data.Second, for the problem of passive localization in acoustic network, thisdissertation presents a target location algorithm based on pseudo-linear least squaresand probability data association. It combines pseudo-linear least squares estimationand probability data association algorithm to locate and track target under the denseclutter.Third, considering the real-time realization in engineering applications, a targetlocalization algorithm is proposed, which is easy to realize in the engineering. It usesthe minimum time difference technique and cross localization algorithm to deal withtime-delay with pre-processing data. Finally, it adopts a reverse validation method toidentify and remove the fake point, which leads to a better fusion results.Fourth, pro-processing algorithm is tested with engineering data. The results showthat localization algorithm based on the minimum time difference technology isfeasibility in engineering application. Finally, some definition of performance evaluation indexes are given, and been verified with engineering data.