Research On The Application Of BP-Neural Network In The Hydrologic Data

Fresh water is a kind of scarce resource in modern society. The distribution of fresh water all over the world is uneven and with the development of the population, the demand of fresh water also has sharply increased, people will face more difficult in use of fresh water. When the challenge is under way, in some sense, we must handle with the contradiction between the highly growth demand of fresh water and sustainable development of the use of nature resource.Research in the Hydrological Sciences has reach a high level, and scientists focus on two branch of this field, the prediction of water's level and flow. In order to get more certification to prove the decision about flood protection, or the management of water resource, we should use more scientific methods in the field of Hydrological research. Many scientists have dedicated to this field and get a lot of achievement.Nowadays, with the development of the Neural Network, more and more researchers begin to use this method in the field of Hydrological research. Using Neural Networks can fit any shape of curves, in consider of the nonlinear feature of Hydrological data, this paper use the Back Propagation Neural Network, so called BP-Neural Network, to do some research.The data set used is Hydrological data from Shaanxi province. We build different model in base of those data, those has different impact factor or different training set. When comparing the predictions of those models, we can receive some conclusions.