A Field Feature Extraction Method of Sand-dust Storm Ensemble Forecast System Based on ANN
Yufeng Yang · Jisuanji fangzhen · 2007
The sand-dust storm happened more and more frequently in resent years. It not only seriously affects the normal life of people, but also do certain harm to the social economy and the environment. According to the origin and characteristic of sand - dust storm weather, a method of field feature extraction of sand - dust storm ensemble forecast system on GA - neural network is proposed in this paper, which is based on the analysis of the shortcomings of existing feature extraction methods. Firstly, the origin physics field sample of the sand -dust storm can be acquired after the preprocessing of characteristic data. Then the field feature extraction is performed by the presented method. On this basis, the dust storm forecast system on GA-neural network is constructed and realized with the aid of the ensemble forecast theory. The emulational results show that the presented method obtains a better achievement in enhancing the veracity of the sand-dust storm forecast.