Application Research On Time Series Analysis For Power Load Forecasting

During recent years, the installed capacity of wind power generation has been developing rapidly. And for its wave and intermittence characters, the influence when it is extensively connected to the electricity grid is no longer able to be ignored. The traditional electrical load forecasting method can not fit the new situation. After checking and reading lots of documents and files, I tried my best to find a better way to adapt to forecast the load of the wind power connected electrical grid.This paper analyzed the factors which influenced the electrical load, and emphasized the influence caused by wind power combined to the grid. When large-scale wind power is connected to the grid, the nonlinearity of the electrical system load will be signalized, it will be more difficult to forecast the load by traditional way. This paper analyzed the characters of the power load, studied the traditional forecasting theories. Most important of all, this paper studied the time series analysis methods on power load forecasting, including the thread of modeling, the suitability checking of the model and the parametric recognition.This paper studied the traditional time series analysis power load forecasting on both theories and application. And I discovered the classical methods are no longer suitable when the wind power extensively connected to the grid. So, this paper uses the steps-conformed methods to estimate the parameters of the model, which can get better definition and constringency. And this paper uses the new weighing analysis methods to optimize the model of the power load forecasting, which still has better definition when the wind power extensively connected to the grid comparing with the classic methods.