Research On Prediction Of The Price About Uranium Resource Product Based On EMD-BP Network Method

Nuclear power has some greater environmental advantages than other traditional energy sources. With the continuous development of society, nuclear power as a clean and economical energy will get more deeply and widely used.The uranium as the material basis for nuclear power, it is important to the development of nuclear power. Along with the continuous expansion of nuclear power development in our country, domestic uranium reserves will not to met the requirements in future. Under this background, the demand for uranium resources in our country will become more and more dependent on the international market,so the research of prices on uranium resource product to our country in the international market have important reference value.In this paper, based on the fluctuation of uranium resource’s price, and after the EMD method decomposed data signal, the BP neural network method is applied to predict the short-term price. This article focus on the prediction of uranium product price and explore the scientific prediction method. It mainly includes three parts, there are six chapters.The first part of the theoretical part, mainly distributed in the first, second and third chapters. This part introduces the time series prediction methods, EMD(empirical mode decomposition method) and BP neural network.The second part of the empirical part, including the fourth to five chapter,respectively made three empirical analysis. The first empirical method is using empirical mode decomposition to study the fluctuation characteristics of natural uranium prices. The second analysis is the traditional linear regression model fitting and forecast adaptability test, for the EMD- BP neural network prediction method to provide a scientific comparison. Finally analysis is based on the EMD-BP neural network method do price prediction.The third part is the conclusion part, it mainly in chapter six.And it come to the conclusion and prospect of the study based on the results of empirical research.In this paper, the innovation points as follows:( 1) It was the first time Combined the EMD method and BP neural network method to forecast the price of uranium research in China.(2) This method makes the prediction accuracy than simply using the modern intelligent methods to forecast.