Research Of Detection-Tracking-Integration Algorithm With Knowledge-Aided

Serving radar generally uses detection before tracking (DBT) system, which will appear a lot of false alarms and missed alarms when facing stealth aircraft or low-altitude small target because of faint echo energy and interference of strong clutter in the surveillance scene. Thus results in significant loss of echo information and makes the detection and tracking performance deterioration. Detection-tracking-integration algorithm with knowledge-aided is designed based on the traditional DBT, which combines with target and environment prior knowledge, associates the detection process with tracking process, and quantizes the relationship between detection and tracking, makes the detection process and tracking process coupled, finally improves the the radar detection and tracking performance for dim targets. As a hot research point for a long time, there are still some technical issues that need to improve and resolve, for example, quantizing the detection and tracking performance in strong clutter environment, coupling multi-target detection and tracking, etc.Regarding to the above issues, the main work of this thesis are as follows:1、A tracking performance prediction algorithm based on modified Riccati equation (MRE) is studied, This algorithm gives a quantitative relationship between the detector operating point and the tracking performance, then provides a theoretical foundation to determine the detector operating point, when the detector works on the optimized operating point, the tracking performance is improved effectively.2、Aiming at dim targets detection and tracking in strong clutter, we study the tracking performance prediction algorithm with amplitude information for the case of a target in heavy-tailed clutter, or more precisely K-distributed clutter.This algorithm gives the method for determining the detector operating point in a complex background, compared with the algorithm of no use of amplitude information, this algorithm achieves better tracking performance。3、 An algorithm integration of Bayes detection with PDAF is studied here, which introduces a feedback loop from the tracker to the detector, and reduces the probability of false alarms effectively in detector, meanwhile, improves the tracking accuracy and successful track probability.4、Aiming at detection and tracking for proximity targets, A multi-target detection-tracking-coupled algorithm is proposed, this algorithm uses parallel Bayesian detectors independently for each target detection, inhibites the target interference with each other effectively, and improves the detection and tracking performance.