基于遺傳算法改進(jìn)bp神經(jīng)網(wǎng)絡(luò)的短期風(fēng)電功率預(yù)測(cè)研究

1、2012年“挑戰(zhàn)杯”大學(xué)生年“挑戰(zhàn)杯”大學(xué)生課外學(xué)術(shù)科技作品競(jìng)賽及創(chuàng)業(yè)設(shè)計(jì)大賽課外學(xué)術(shù)科技作品競(jìng)賽及創(chuàng)業(yè)設(shè)計(jì)大賽作品名稱：基于遺傳算法改進(jìn)BP神經(jīng)網(wǎng)絡(luò)的短期風(fēng)電功率預(yù)測(cè)研究作品類別：社會(huì)科學(xué)類項(xiàng)目成員：劉知發(fā)陳軍母桑妮向亞軍唐艷利聯(lián)系電話：15196009869完成時(shí)間：2012年3月20日ABSTRACTWindpowergenerationoneofthemostsignificantresearchsubjectsin21stce

2、nturytopsamongthecleanrenewableresourceswhichplaysacriticalroleinreducingpollutioneasingstressproducedbyenergyconsumption.ThethesisusingmethodsincludingtimeseriesgeicalgithmBPnerveswkestablishes4predictionmodelswithwindp

3、ower.ThroughMatlabprogrammingtheautharrivesatvariouspredictionresultscheckstheiraccuracy.Firstofallthethesisanalysesdevelopmentsituationofwindpowergenerationathomeabroad.onthebasisofthefirststepthesecondchapterdiscussest

4、heissuebymobileaveragepredictionmethodromtimeseriespredictionmethodBPnerveswkpredictionmethod.ThroughmixedprogrammingofExcelMatlabtheaccuracyrateofaveragepredictionmethodromtimeseriespredictionmethodBPnerveswkpredictionm

5、ethodis82%70%84%respectivelywhilequalifiedrateofthemis85%65%92%.ApartfromtheabovementionedtheauthalsodrawsthatBPnerveswkpredictionmethodissuperitotheothertwoones.ThenusingthedatafecastbyBPnerveswkpredictionmethodtheauthm

6、akingpredicativelycomparativeerranalysisacquiressixgroupsofcomparativeerrsofpredictionfigure(seet1)drawsananalysistofcomparativeerrs.Asumupthatgatheringofwindturbineisabletodecreasepredictiveerrsofwindpowerisreachedbymea

7、nsoferranalysis.Thecausesgivingrisetosuchaconclusionareinterpretedaswellanideathatincreasinggatheringdensityofwindturbinecanreduceexpectantjudgmentoferrsisalsoputfward.BasedontheChapter2amodelcombingwithgeicalgithmmethod

8、BPnerveswkisestablishedtofecastwindpowersoastoincreasinglyimproverealtimemonitingaccuracyofwindpower.Withthemodeltheauthmakingaweeklongpredictiontowardsanthwestpowerplantcomparingwithactualfigurescomestoaconclusionthatth

9、eaccuracyqualifiedratesareupto89%95%.Ultimatelytheauthanalyzestheconclusionofthesubjectsummarizesthepredictiveresultsexistingproblemsproposessomeaspectsthatdemimproving.KeyWds:predictionofwindpowerromtimeseriesBPnerveswk