搜索
1电力系统短期负荷预测目录中文摘要:..............................................................3英文摘要:..............................................................41绪论..................................................................41.1短期负荷预测的目的和意义.......................................51.2电力系统负荷预测的特点和基本原理................................61.2.1电力负荷预测的特点........................................61.2.2电力负荷预测的基本原理....................................61.3国内外研究的现状...............................................71.3.1传统负荷预测方法.........................................81.3.2现代负荷预测方法.........................................81.4神经网络应用于短期负荷预报的现状..............................101.5本文的主要工作................................................102最小二乘法...........................................................122.1最小二乘法原理................................................122.2多项式拟合具体算法............................................122.3多项式拟合的步骤...............................................132.4电力系统短期负荷预测误差......................................142.4.1误差产生的原因..........................................142.4.2误差表示和分析方法......................................142.4.3拟合精度分析............................................153基于神经网络的短期负荷预测...........................................173.1人工神经网络..................................................173.1.1人工神经网络的基本特点..................................173.2BP网络的原理、结构............................................173.2.1网络基本原理.............................................173.2.2BP神经网络的模型和结构..................................183.2.3BP网络的学习规则........................................183.3BP算法的数学描述..............................................193.3.1信息的正向传递...........................................193.3.2利用梯度下降法求权值变化及误差的反向传播................193.4BP网络学习具体步骤............................................203.5标准BP神经网络模型的建立.....................................2123.5.1输入输出变量............................................213.5.2网络结构的确定..........................................213.5.3传输函数................................................223.5.4初始权值的选取..........................................233.5.5学习数率................................................243.5.6预测前、后数据的归一化处理..............................243.6附加动量的BP神经网络.........................................243.6.1标准BP算法的限制与不足.................................243.6.2附加动量法..............................................254算例分析.............................................................274.1负荷数据......................................................274.1.114天实际的负荷数据......................................274.1.2归一化后的负荷数据......................................294.2两个模型仿真后的结果分析......................................324.3两种模型拟合精度分析..........................................394.4附加动量法....................................................41结论..................................................................42谢辞..................................................................43参考文献..............................................................44附录1最小二乘法的MATLAB程序.........................................46附录2标准BP神经网络的MATLAB程序....................................48附录3附加动量法的MATLAB程序.........................................513电力系统短期负荷预测摘要:电力系统负荷预测是电力生产部门的重要工作之一。准确的负荷预测，可以合理安排机组启停，减少备用容量，合理安排检修计划及降低发电成本等。准确的预测，特别是短期负荷预测对提高电力经营主体的运行效益有直接的作用，对电力系统控制、运行和计划都有重要意义。因此，针对不同场合需要寻求有效的负荷预测方法来提高预测精度。本文采用神经网络方法对电力系统短期负荷进行预测。本文主要介绍了电力负荷预测的主要方法和神经网络的原理、结构，分析了反向传播算法，建立三层人工神经网络模型进行负荷预测，并编写相关程序。与此同时采用最小二乘法进行对比，通过对最小二乘法多项式拟合原理的学习，建立模型编写相关程序。通过算例对两种模型绝对误差、相对误差、拟合精度进行分析，同时比较它们训练时间，得出标准BP神经网络具有更好的精度优势但训练速度较慢。最后针对标准BP神经网络训练速度慢、容易陷入局部最小值等缺点，对标准BP神经网络程序运用附加动量法进行修改，分析改进后网络的优点。关键词:短期负荷预测,标准BP神经网络，最小二乘法，附加动量法4TheShort-TermLoadForecastingofthepowersystemAbstract：Powersystemloadforecastingisoneofthemostimportantworkoftheelectricityproductionsector.Theaccurateloadforecastingcanarrangeunitstart-stop,reducethesparecapacity,reasonablearrangementofthemaintenanceplanandreducepowercost,etc.Ithasadirecteffectontherunningefficiencyofthepowermanagemententitiesandalsohastheimportantmeaninginthepowersystemcontrol,operationandplanning.Soitisimportanttofindeffectivemethodtoenhanceforecastprecisionfordifferentoccasions.Inthispapertheneuralnetworkisusedfortheshort-termloadforecastingofthepowersystem.Thisarticleintroducesthemethodofthepowerloadforecastingandtheprinciples,structure,back-propagationalgorithmoftheneuralnetwork.Thenthethree-layerartificialneuralnetworkmodeliscreatedforloadforecastingandtheprogramiswritten.Atthesametime,theleastsquaremethodisusedforcomparing.Bylearningthepolynomialfittingprincipleofthesquaremethod,themodeliscreatedandtheprogramiswritten.Throughcomparingtheabsoluteerror,therelativeerror,thefittingprecisionandtheirtrainingtimeofthetwomodels,theBPneuralnetworkisprovedtohavebetteraccuracybutslowertrainingspeed.DuetothestandardBPneuralnetworkhasslowertrainingspeed,easytofallintothelocalminimumvalueandothershortcoming,theadditionalmomentummethodisusedtomodifythestandardBPneuralnetworkandtheadvantageoftheimprovednetworkisconcluded.Keywords：Short-termloadfo