Research On Wind Power Forecasting And Unit Dispatch Of AGC

With increasing of coal, oil and natural gas reserves dwindling and carbon dioxide and other greenhouse gases. The use of new energy is receiving more attention. Wind energy as a kind of clean, abundant reserves renewable energy is the most important energy of new energy source。Due to the random characteristic of wind speed, wind electricity generating system uncertain factors, this pair of power between the economic operations of the new challenges. How to reduce the instability of wind power, to protect the security and stability of power grids, it requires the wind speed and wind power to be predicted. How to ensure power grid load requirements based on wind power prediction, AGC unite must call on to distribute the power system load. But in a pledge to meet regulatory requirements under the premise of which called units, the number of call-conditioning capacity, in order to achieve the best economic purposes, This requires the use of some methods to calculate the most optimal way.First, use artificial neural network to build the wind speed prediction model based on support vector machine. Taking into account the seasonal characteristics and highly non-linear of wind speed. Therefore, take the wind speed data in different years but the same day as an important parameter, to do wind speed forecastingSecondly, the gray model for this article was chosen to do wind speed short-term forecasting, Due to the limitations of the gray model GM(1,1), there is a big error between the results and actual value. For Gray model GM(1,1) for wind speedforecasting can not be effectively used to reflect the seasonal changes of wind speed and low forecasting accuracy problem, Promote the model GM(1,1) to Model GM(1,1,λ),And use differential evolution algorithm for solving the optimal solution model. . Compare the results with the gray model forecasting results.Then, calculate the wind power with the wind speed forecasting data combined the wind turbine power curve together. The results will be used for modeling of wind farm output power, to provide data to support AGC units deployment for below text.Finally, this paper discussed the load distribution of AGC units, Calculate the AGC regulation capacity of power system based on the predictive value of wind farm output power. Establish the function model which is take lowest cost of AGC regulate fee as target based on calculation AGC regulation capacity of power system, solve the model with genetic algorithms. It will receive the start and cease, load distribution and minimum deployment cost of AGC.