Target motion analysis and track association with a network of proximity sensors

摘要

Target motion analysis and track association are the aims of this paper. It is assumed that the target trajectory is only partially observable by using temporal processing of a single sensor. Thus, original algorithms have been developed for fusing local estimates. Though suboptimal, such algorithms are feasible and close to optimality. Their performances have been theoretically investigated. Another problem is track-to-track association, the aim of which is to perform decision if the tracks estimated by two local subsystems are related to a single target or not and for which an original approach has been developed. This study is illustrated by a specific case study: the detection of a moving target by stationary electric sensors.