資源描述：
《Application of an improved adaptive chaos prediction model in aero-engine performance parameters英文資料》由會員上傳分享，免費在線閱讀，更多相關內(nèi)容在學術論文-天天文庫。
1、WSEASTRANSACTIONSonMATHEMATICSChunxiaoZhang,JunjieYueApplicationofanimprovedadaptivechaospredictionmodelinaero-engineperformanceparametersCHUNXIAOZHANGJUNJIEYUECollegeofScienceCollegeofAeronauticalEngineeringCivilAviationUniversityofChinaCivilAviationUniversityofChinaTianjinTianjinCHINACHINAc
2、xzhang@cauc.edu.cnjjyueyjs09@cauc.edu.cnAbstract:Basedontheresearchofcomplexityandnon-linearityofaero-engineexhaustgastemperature(EGT)system,aregularizationadaptivechaoticpredictionmodelappliedinshorttimeforecastingofEGTwasproposed.Inthisresearch,wedevelopanewhybridparticleswarmoptimization(H
3、PSO)arithmeticinordertoimprovetheaccuracyoftheforecastingmodel.Thisarithmeticenhancedtheabilityofdealingwithintegervariablesandconstraintsbyaddingandchangingsomemanipulationsto?tinwithoptimizingcontinuousandintegervariables.ThetestresultsarebasedonQARdatasuppliedbyacivilairlinecompany,andshow
4、thattheproposedframeworkperformsbetterthanthetraditionalchaoticforecastingmodelonpredictionaccuracy.Therefore,thisarithmeticisef?cientandfeasibleforashort-termpredictionofaero-engineexhaustgastemperature.KeyWords:Exhaustgastemperature(EGT);Regularization;Adaptivechaosprediction;Hybridparticle
5、swarmoptimization(HPSO);Principalcomponentregression(PCR);Aero-engine.1Introductionleastsquares(OLS),themodelexistedtheseriousmulticollinearitythatwillincreasethepredictinger-Advancedmonitorandprognosticschemestodeter-ror.Someresearchersandscholarsforecastedenginemineengineconditionisimportan
6、tformoderncivilsystemsreliabilitybyneuralnetworkmode[5],theyaircraftinordertoreduceunnecessarymaintenanceimprovedBPneuralnetworktodynamicallyforecastactionandimproveaircraftsafety.Thus,thecur-read-time,buttheoverlearningandunstabletrainingrentestimatesofengineconditionarenecessarybe-ofneuraln
7、etworkwereinsurmountableproblems.Inforeupcoming?ghtstoavoiddelays.Infact,ab-[6],theKalman?ltermethodwasproposedtoesti-normalityofenginecanbeidenti?edbymonitoringmatetheaero-enginehealthparameters.Kalman?lterperformanceparametersofengine.There
