简介:ChinaGuangdongNuclearPowerGroup,ChinaPowerInvestmentGroupandGuangxiInvestmentGroupreachedaframeagreementtojointlybuildthephaseIprojectofGuangxiBailongNuclearPowerPlantinJuly.Accordingtotheplan,GuangxiBailongNuclearPowerPlantwillhaveatotalinstalledcapacityof6000MW.ThescaleofitsphaseIprojectisplannedtoreach
简介:Accuratepredictionofwindpowerissignificantforpowersystemdispatchingaswellassafeandstableoperation.BymeansofBPneuralnetwork,radialbasisfunctionneuralnetworkandsupportvectormachine,anewcombinedmethodofwindpowerpredictionbasedoncooperativegametheoryisproposed.Inthemethod,everysingleforecastingmodelisregardedasamemberofthecooperativegames,andthesumofsquareerrorofcombinationforecastingistakenastheresultofcooperation.TheresultisdividedamongthemembersaccordingtoShapleyvalues,andthenweightsofcombinationforecastingcanbeobtained.Applicationresultsinanactualwindfarmshowthattheproposedmethodcaneffectivelyimprovepredictionprecision.
简介:超短期风速预测对风电场功率预报系统的建立和运行至关重要。针对具有较大随机波动性的风速预测,研究了一种基于误差修正的极端学习机(ELM)超短期风速预测方法。利用ELM模型对风速进行初步预测,并利用由此得到的误差数据样本建立差分自回归滑动平均模型(ARIMA),进行误差预测,最后使用预测误差对风速的初步预测值进行补正,从而求得最终预测值。仿真实验结果表明,该方法在风速超短期预测中的可行性及有效性。