Airborne Fire Control Radar Target Tracking Technology

The multi target tracking problem of airborne fire control radar involving different techniques is studied in this dissertation. The techniques of data association, tracking, filtering and prediction of targets are studied. The multi-target information, such as their radial distance, movement parameters are received from radar signal process. After getting the information, the multi-target tracking problem can be expressed as a series of operations, including state estimation and update. Finally, real-time multi-target tracking will be completed with high precision and stable tracking traces will be formulated.The main works of this dissertation are as follows:1,The model of radar data process in the project of"digital radar simulation system"are introduced. By processing its sub-models of data association, track filtering, identification and threat assessment, the trace management, target identification and threat assessment are completed, then, all the information are transferred to the radar control central and radar display model for displaying.2,The state estimation and filtering methods are studied. This dissertation focuses on the application of nonlinear filters. They are based on linear filters and used in nonlinear situation with nonlinear state and/or measurement equations. In this dissertation, three kinds of nonlinear filters are studied according to different environment. They are referred as extended Kalman filter, unscented Kalman filter and particle filter. Their performances are compared in simulation experiments. The results help to apply them in different environment. So when applying them in various and different environment could get a best estimating precision meanwhile ensuring the tracking efficiency.3,Data association methods are studied. The relationships of the measurements or traces are set up among different frames. It decides whether these various measurements are from the same target. Two kinds of data association algorithms are studied in the dissertation. They are referred as nearest neighbor data association and joint probability data association. The former can be utilized with sparse targets or clatters, and the latter used in the opposite situation with dense targets or clatters. Their performances are compared in simulation experiments.4,The algorithms of multi target tracking initiation and termination methods are studied. They are two of the fundamental parts in multi target tracking theory, and they are absolutely necessary to set up new and delete redundant traces. In this dissertation, methods used in sparse and dense clatters are studied respectively. Simulation experiments are given in the dissertation.In the dissertation, we pay emphasis on discussing the problems of state estimation,filtering,tracking, data association and so on. Some progress has been made. Nowadays, in the area of airborne fire control system, multi target tracking has to cope with complicated environment and dense objects, so more and more studies are supposed to do deeply.