城市燃氣輸配調(diào)度系統(tǒng)的研究

1、遼寧科技大學碩士學位論文城市燃氣輸配調(diào)度系統(tǒng)的研究姓名：梁金鳳申請學位級別：碩士專業(yè)：化學工程與工藝指導教師：馬慶元20070301遼寧科技大學碩士論文AbstractShorttermgasloadforecastingistheimportantpartofthegastransmissionanddistributionsystemThemostimportantsignificanceisthatpeoplecandefinet

2、hebestinletgasplanandpeakloadmethodwithit，anditcanhelppeopleguaranteethegasloadandoptimizeddispatchingofthepipenetworkInthistext，theauthorcombinethewaveletmethod，dataexcavationandartificialneuralnetworkthatnewresearchmet

3、hodtoforecastshorttermgasloadWiththewaveletpackanalysismethod，theauthordenoisethehistoricaldata，Zeroequalizingvaluenormalizationfromwhichdataexcavationisnotonlyguaranteeingtheprecisionofthestudyresultofneuralnetwork，buta

4、lsohelpingtoincreasethestudyspeedofneuralnetworkTheauthorusethreeimprovedBPneuralnetworkforecastedmodelthatestablishedtoforecastonecitydailygasloadAmongthethreemodels，theimprovedBPneuralnetworkforecastedmodelwithinputtin

5、gmomentumselfadaptingalgorithmhashighprecisionandspeedoffastconvergenceTheimprovedBPneuralnetworkforecastingmodelshashighprecision，butthenetworkstructureisverycomplex，hardpopular，Toovercometheseshortcoming，theauthoruseth

6、eforecastingmodelofRadialBasisFunctionneuralnetworktoforecastonecitydailygasloadin2002，theaveragerelativeerroris402％inFebruary，theaveragerelativeerroris694％inMarchWiththesamemodel，theauthorforecastanothercitydailygasload

7、in2004，theaveragerelativeerroris311％inFebruary，theaveragerelativeerroris199％inMarchTheforecastingmodelofRadialBasisFunctionneuralnetworkissettlingthequestionofplannedlevelnumberandneuralelementnumber，andtheforecastingpre