The Forest Fire Size Class Forecasting System On BP Neural Net

In the application of Geographic Information System, many applications are used in the special theme, but in the real application, it is popular that every theme goes his own way. Because of the barriers between higher and lower levels or between different departments and regions,They don't share the data built in the application systems, city management systems and involving departments under the central leadership, so the cost is very high. More fore, each application system is developed by different development system, so the data could not be supported by others. The digit islet is becoming real. In the traditional database, the database is only supported relation database popularly. The GIS database has not only attribute data but also spatial data. The special data decides the relations in the location, attribute and surface features,so the databases in GIS show the different ways accessing to the databases. The application system is able to get the attribute data and the special data. The special data has great amount. So the ways access to the database are defined by the sharing calculating tasks in view of Web GIS. The ArcView software is developed by ESRI,which may used by customs with Visual basic and other development software together to develop the application software displaying the interesting map. In network, the GIS software may work accompanying to the program developed by the users showing the map. Forest is playing an the more important role in economy accompanying the growing of economy. Before the BP neural net forecast fire size class, it needs a process of studying from sample data. The neural net adjusts the weight value and threshold value according to the sample so as to give the linking weight value and threshold to low the difference between output from itself and the expected value. The net must have the ability in the memory and the association through much practice from sample data.The forest fire size class forecasting may success through many times practices if the atmosphere circulating current index and fire index are known. In the Web, subscribers may forecast the class of the forest fire size using the program in the local database and get the interesting regional view which is shown by the GIS software.