《Artificial Neural Network Ability in Evaluation of Random WaveInduced Inline Force on A Vertical Cylinder》.pdfVIP

ChinaOcean Eng．，V01．26，No．1，PP．19—36 Ocean ◎2012Chinese and Berlin EngineeringSocietySpringer-VerlagHeidelberg DOI 0890-5487 10．1007／s13344—012—0002—8，ISSN ArtificialNeuralNetworkinEvaluationofRandom Ability InlineForceonAVertical Cylinder Lotfollahi—Yaghin，M．A．3”，Pourtaghi，A．5， Sanaaty，B．。and Lotfollahi—Yaghin，A．8 8 FacultyofCivilEngineering,UniversityofTabr拓，Tabriz，Iran o Researchers Azad Young Club，ArdabilBranch，IslamicUniversity，Ardabil,Iran andOcean 。DepartmentofNavalArchitectureEngineering,OsakaUniversity，Japan 12 revisedform6 15November (ReceivedMay2011；received September2011；accepted 2011) ABSTRACT An basedonartificialneural isusedto relationsbetween approach network(ANNldeveloppredictive inlineforceonavertical andsomeeffective datausedtocalibrateandvaiidate hydrodynamic cylinder parameters．The are neuralnetworksthatarctrainedwith theANNmodelsobtainedfroman feed．forward the experiment．Multilayer areconstructeduseofthree surface and algorithm by designparameters(i．e．waveheight，horizontal back-propagation vertical networkandtheultimateinlineforceasthe isconducted velocities)asinputs onlyoutput．Asensitivityanalysis ontheANNmodelsto the the