風電短期負荷預測中的若干基本問題.pdf

1、天津大學碩士學位論文風電短期負荷預測中的若干基本問題姓名：鄒淑春申請學位級別：碩士專業(yè)：控制科學與工程指導教師：楊正瓴201112ABSTRACTTheresearchofshorttermloadforecastingismoreandmoreimportantespeciallyinelectricpowerSeparatingwhitenoisefromshorttimeloadseries，andreducingforecast

2、ingerrorsaremoresignificantforthesaferunningofthepowersystemandtheeconomicdevelopmentThewhitenoiseseriesmakeanupperlimitofloadforecastingaccuracyandisanessentialdataforfindingbetterforecastingmodelsIntherealcomplexsystem

3、s，temperature，andthefinitedigitalcomputercanproduceadditiveerrors，reduceforecastingaccuracyHowtosolvethoseproblemsisthestudyofthisarticleThemaincontentsandresultsare：(1)AccordingtotheCram6r’Sdecompositiontheorem，theloadt

4、imeseriescanbedecomposedsignalandnoiseWhentheloadtimeserieshavethefrequencysignalandhighfrequencynoise，thevarianceofwhitenoisecanbeestimatedbyusingthemethodofdifference(2)Basedonthemethodofwavelet，thetimeloadseriescanbee

5、asilydecomposedaseriesofsignalandnoise，butitishardtoidentifywhichisthebestwhitenoiseUsethetheoryofStatisticsMethod，wecanfindthebestwhitenoiseThisarticlestudymorenumbericaltests(3)AccordingtotheShannonsamplingtheorem，thet

6、imeloadserieswhichweresampledfromcomplexsystemsintherealworldalmosthavewhitenoiseserieswithinthemFortheweathersensitiveloads，theShannonsamplingtheorempredicatesthatthetraditional‘‘day’’samplingmethodswhichwereusedtoelimi

7、natethefluctuant／non—stationaryofloadseriescanusuallyenlargethewhitenoisesProposethemethodoflowordernonlineartransformationsteducetheforecastingerrors(4)Employthesamepredictionmodeltotesttheperformanceofpowerfunctiontran

8、sformationandlogarithmictransformation(alowordernonlineartransformation)ThisarticlestudymorenumericaltestsKEYWORDS：Shorttimeloadforecasting，WhiteNoise，LowOrderNonlinearTransformations，PowerFunctionTransformation，Logarith