The Statistical Analysis Of Temperature And Short-term Forecast

This thesis mainly carried on the statistical analysis to the temperature in Beijing over the past decade by different methods and arrived at some different statistical models, then utilize the resulting statistical model to calculate predicted values and compare with the real values .Finally,we can have a statistical model related to a particular month,which has the smallest relative error. Moreovre,we can utilize the statistical model to do a short-term forecast. This thesis gives time series of the temperature in Beijing, statistical simulation and forecasting functions, This method is suitable similarly for the electricity fluctuations and other fluctuations. In addition, the thesis also utilize interval estimation and analysis of variance to do the temperature estimates and projections,and then have a number of specific prediction equations. This thesis can be divided into three parts:The first part describes the research background and research methods of this article. in addition to,there is a brief overview of two important ways in mathematical statistics: interval estimation and analysis of variance.The second part mainly utilizes time series analytic method to carry on the analysis and forecast of the temperature . In this part,we mainly utilizes the straight-line trend of law and the trend method of the quadratic regression curve to analyze the data and come to the corresponding predicted formula.The third part mainly analyzes these data by the linear regression method, gray prediction theory and methods such as cluster analysis and obtains predicted equation, and then utilizes the predicted formula to calculate the predicted value.Finally we can take the formula which has the smallest relative error as the best prediction models.