Research On Methods Of Multiple Passive Sensors Measurement Data Association

The techniques of target tracking and data association are important topics of the research on multiple passive sensors information fusion system. Because they have been found wide applications in both military and civil areas, many countries and researchers have paid much attention to the development of the target tracking and data association technology. However, there are many problems in the passive system, such as the nonlinear measurement, lack of distance information, low signal-to-noise ratio and high false alarm rate, which will bring great difficulty with target tracking. Aiming at the problems above, this dissertation mainly discuss the research of data association for target tracking based on multiple passive sensors and it's proposed some real-time and effective new method.Firstly, it's briefly described the background, significance and present situation of multiple passive sensors information fusion system, on the base of which, it's studied data association in two aspects, measurements-to-tracks and measurements-to-measurements, and showed flow, relative merits and application scenarios of each algorithms.Secondly, analyzed the shortage of the method based on the distance of line-of-sight, it's proposed a data association algorithms based on the cotangent of angles in multiple passive sensors, which makes use of angle information directly, constructs statistical quantity reasonable, screened by azimuth angle test and elevation angle test, and then the indicated function analysis is used to select the correct association from the set of candidates. The proposed algorithm is much faster than the conventional ones without the decrease of performance.Thirdly, for the selection to cost function in passive multi-targets data association, it's proposed a modified cost function in passive system. Comparing with conventional function based on generalized likelihood ratio, the new one considers estimate error in fused target position, which is more accurate to reflect the relationship between measurements and targets. It's showed good results in the application to Lagrange relaxation algorithm.Finally, according to the study and comparison of least square to multiple passive sensors system location problem, a location algorithm based on distance weighted least squares is proposed, which makes use of the information of pseudo-measurements for weighting , and improves the correct association with lower computing expense.