Gray Modeling Theory And Its Application In Target Recognition

The concept, characteristic and field of application of gray model theory and the property of gray model are introduced systematically in this paper. The mechanics and The method of building GM (1, 1) and GM (1, N) are deeply researched and discussed. The requirements and relative questions, which should be paid attention to when building gray model, are also discussed.Because it needs a lot of fund and time to obtain the real data of target noise, there is limited quantity of real data now. In this paper, on the basis of deeply analyzing the spectrum feature of the target noise, a large quantity data of two sorts of target noise are simulated under the condition of different frequency, departure, depth and SNR.The applications of GM (1,1) and GM (1, N) in target recognition are individually researched on the basis of theory analysis. The scheme of target recognition based on gray model is set up. GM (1, 1) and GM (1, N) of target noise are also set up. The method of target recognition, which takes gray model parameter as features, is put forward after their models and the characteristic of model parameter are deeply analyzed and researched. The availability and feasibility of the target recognition method based on the gray model are researched and verified with simulated and real measured data, associative neural networks. The results of target recognition based on GM (1,1) and GM (1, N) is compared finally.The research in this paper combines closely the requirement of modern target recognition, represents the characteristic of research and application of new method. It has the higher value of theory and actual for improving accuracy of target recognition.