Track Fusion For Bistatic High Frequence Surface Wave Radar

Bistatic radar system not only in theory but also in practice has been approaching maturity. Although the research of high-frequency radar started not early, it has a made great achievement. However, the research of dual / multi-frequency radar base still remain in the initial stage, especially that the research in the data fusion of dual / multi-frequency radar mostly remain in the theoretical study and simulation analysis.With the bistatic high frequency surface wave radar as the background, this paper first did the analysis of the characteristics of real data, including the main error source of coordinate transformation, conversion method for all parameters and the quality of track formation. Through the analysis of real data and the analysis of the traditional track correlation algorithms, this paper illustrates the unique problems that exist in bistatic high frequency surface wave radar track association , paving the way to propose the new track association algorithm later. Then, the fusion method named Track Fusion Based On Track Segment Method for bistatic high frequency surface wave radar is proposed based on the above considerations. With the track segment as the basic unit , it not only effectively solutes the problem of the periodic decline quality of the track segment but also eases the phenomenon exists in traditional methods which is first appeared before association. The determine guidelines of the length of the track segment, method for solving the integrated correlation and guidelines of track segment association are focused on in this section. Then, the concepts, frameworks and structures of the three-dimensional model combining Neural Network, Fuzzy Reasoning and Expert System are introduced, and then the NFE model is applied to the track fusion of bistatic high frequency surface wave radar system. By using positioning distance difference, radial velocity difference and speed difference, this paper do the judgment of, the track segment correlation, solved the complexity and uncertainty exists in the track correlation.