Library Of Radar Clutter Modeling Technology And Its Application Research

Clutter often has typical heterogeneous in the real world, which will greatly lead to the performance degradation of detecting target by traditional techniques. Now, the real-time prior knowledge about the radar surrounding environment must be exploited, contributing to detect the target based on the matched environment. To obtain the prior knowledge effectively, establishing the knowledge database including target and clutter database is pre-requisite, which is also the most important part of the knowledge aided (KA) system and has significant value of research.This paper focuses on the modeling techniques of clutter database, employing the prior geographical information, such as digital terrain elevation data (DTED), geographic information system (GIS) and so on. And the applications in the radar are given through KA algorithms. It mainly includes:1. For the use and fusion of prior geographical information, a ray-tracing algorithm to analyse terrain visibility with DTED is proposed, the methods of computing terrain slope and depression angle using DTED are given, according to GIS, the methods of classifying terrain and obtaining static clutter information combined with the empirical clutter models are introduced, hard and fuzzy classification to divide the clutter regions which can fuse DTED and GIS are studied.2. For the simulation of correlated non-Gaussian clutter, the basic principles of Zero Memory Nonlinearity and Spherically Invariant Random Process methods are discussed, the steps of the algorithms are given, the estimations of the parameters of four clutter distributions like Rayleigh are derived, the advantages and disadvantages of these two algorithms are compared, and how to choose them in practice is recommended.3. For the strong and discrete clutter generating by isolated scatters, the algorithms to remove them are discussed, which include GIS selector and the data adaptive selector, and via the KA generalized likelihood ratio detector, the validities of the two selectors are proved by simulation. 4. For the existing problems of the traditional techniques to detect target in the real world, the recursive least square algorithm based on KA is considered, which improves the performance of clutter and noise suppression in the complex environment. Two types of KA-CFAR detectors are discussed, having certain advantages over the numbers of false alarms and performances of detection demonstrated by real data. A kind of KA tracker is proposed, which significantly reduces the numbers of false tracks.The effectiveness of the above work has been verified through simulation experiments. The results show that the modeling techniques of clutter database can aid the radar algorithms well and detect the target effectively.