The Study Of Indoor Target Tracking

As a kind of radar data processing technology, target tracking is widely used in various types of radar system in military and civilian field from the initial single-target tracking to current multi-target tracking technology, and its problems have become increasingly complex. Through-the-wall detection radar as an electronic device can penetrate non-metallic wall to detect people, and due to the irradiated scene normally closed, multi-path propagation phenomenon of electromagnetic wave is obvious and then the detector will detect multipath-false targets. With the help of classical theory of radar data processing, and comprehensive consideration near-field application environment of radar and the target's movement characteristics, we design a kind of indoor target tracking system under the condition of existing multipath-false targets in the paper. It has to keep track of true targets and suppress multipath-false targets, and not only track the targets with uniform motion in a straight line accurately, but also adapt to the fast maneuvering.This thesis describes the indoor target tracking system; the primary work is as follows:First, analysis the deficiency of current statistical model(CSM) algorithm for indoor targets tracking, we put forward a current statistical model algorithm based on maneuvering detection. The simulation results demonstrate that on one hand it can overcome the KF problem that filtering divergence caused by prediction residual error increase while target is maneuvering, but prediction covariance and filter gain not changing with residual error synchronously; on the other hand make up for target tracking accuracy of linear motion for CSM. The tracking ability is between KF and CSM algorithm.Second, Build the model of spread of electromagnetic wave in the indoor, and write the mathematical expressions to estimate multipath propagation delay, and describe the numerical estimation method to solve the coordinate of multipath-false targets while target and radar are separated by a wall or not. The simulation results verify correctness of the numerical estimation method.Third, Put a track as a node, the relationship between track and track as an edge, and use the mathematical model of complex network to describe the indoor target tracking system, and introduce the multipath suppression algorithm based on node degree and based on node coupling, which decide which node in network stands for the true target through threshold judgment criterion. Simulation experiments show that the two types of multipath suppression algorithm can suppress multipath track well, and the multipath suppression algorithm based on node coupling can overcome the problem that a representing true target node in another node's associated wave gate causes the multipath suppression algorithm based on node degree failure easily.