Research On Neural Network BP Algorithm And The Application Of Meteorological Operations

Typhoon is a serious nature disaster which can result in a big number of personnel's dead and property loss in a short period of time Because of its unpredictability. It also can direct or indirect cause other disaster and social disaster of a broad sense. Leading the Typhoon information system to prevention and mitigation field, the static programming can be converted into dynamic programming, this will make due contribution for improving the prevention and mitigation levels of cities and renewing the concepts. With the development of science and technology, The method for studying and analyzing data is improved. So we are eager to seek a better way to manage, analyze typhoon data. Recently, with the Extensive applications of GIS in various fields, Google Map becomes a highly effective approach of data management and analysis.The typhoon information system introduced in this thesis is based on Google Map. Using "Neural net model" as the mathematics base to analyze and forecast the typhoon data. By analyzing and studying the traditional prediction method, this paper designed a new plan which improved the traditional BP algorithm and make some experiments on it. Through comparing and analyzing the experiment results, the excelsior accuracy of typhoon information forecasts was verified, the new algorithm is applied to typhoon forecasting information system, combined with Google API and database technology, making the information more visually. All this make the system widely used and support decision making in defense and repairing for government.This thesis firstly analyzes the significance and the background of this project, and introduces the development,application,study status of Neural net. Then briefly summarized Google API's applications in map. With an introduction of the principle and application of Neural net, advantages and disadvantages of traditional BP algorithm,combining the fact, most of important, proposed improves of the BP algorithm, then analysis and discusses about the comparison between the improved algorithm and BP algorithm, the results show the effectiveness of the new method. Besides, the design consideration, development of the measurement system, architectural pattern, function module design and SQL resolution is described and realized, and then applied the improved algorithm to the system. This paper mainly introduces an improved algorithm based on BP to predict the typhoon data in the typhoon information system with Google API technology, and the improved algorithm has the higher accuracy. Besides, Meteorological information platform based on Google Maps was built as for information interactions and sharing, with the technology of Ajax implemented on the client side, the system has high performance and wide application in practical life.At last, as for the status of this system, we give some suggestions in order to make the system Perfect. Then points out that we make efforts in the direction.