Full-Text Search:
Home|Journal Papers|About CNKI|User Service|FAQ|Contact Us|中文
《Machine Design and Research》 2004-01
Add to Favorite Get Latest Update

Application of the Neural Network Response Surface to Aircraft Preliminary Optimization Design

XING Xiao-nan~1,Xu Yuan-ming~1,Li Suo~1,YANG Xiao-han~2 (1.Flight Vehicle Design Department,Beijing University of Aeronautics and Astronautics,Beijing 100083,China;2.School of Life Science and Technology,Xi'an Jiaotong University,Xi'an 710061,China)  
The concept and mathematical description of the response surface method are presented in this paper.The construction of response surface models by the artificial neural network is analyzed.The response surface can resolve the coupling relationship among the different subspaces in multidisciplinary design optimization.The data exchange and coordination among the subspaces could be completed by the approximation of the neural network response surfaces.The optimization result can be found in the design space by such approximation.This paper describes an example of solving the aircraft conceptual sizing in preliminary design stage.The subspace and the system level all adopt the genetic algorithms as the optimization methods.The results of same problem were compared with those only adopting genetic algorithms without response surface approximations.It showed that the response surface models based on neural network could reduce the number of system analysis and greatly inerease the models accuracies and finally the better scheme could be found out in the design space.raciesand
【Fund】: 北航凡舟科研基金资助项目 (2 0 0 2 0 40 8)
【CateGory Index】: V221
Download(CAJ format) Download(PDF format)
CAJViewer7.0 supports all the CNKI file formats; AdobeReader only supports the PDF format.
©2006 Tsinghua Tongfang Knowledge Network Technology Co., Ltd.(Beijing)(TTKN) All rights reserved