本文以倒傳遞模式之類神經網路來預測彈性支撐梁之彈性支撐剛性大小。考慮梁結構邊界為彈性支撐要比簡支撐和固定支撐之邊界條件來的實際,以往力學分析都是在假設已知彈性支撐剛性值的前提下所作的分析,但在實際的應用上,由於傳統分析方法不容易反向求得彈性支撐剛性大小,故分析結果無法直接提供設計參考之用。本文仍應用類神經網路技術, 將不同彈性支撐剛性所作力學分析之前五個自然頻率的結果,作為網路訓練的樣本,設計出一個能快速地預測彈性支撐剛性大小之倒傳遞神經網路,可直接提供設計參考之用。 A new method based on the technique of artificial neural network for the prediction of the elastic restraint stiffness of beams is presented. The boundary condition of elastic restraints are more practical simulations than the conventional ones, simply-supported or clamped boundary conditions. The conventional structural analysis of beams with elastically restrained boundary is proceeded on the known elastic restraint stiffness. In practice, the elastic restraint stiffness of beams can not be easy evaluated with the conventional analysis. In this paper, a network is trained by backpropagation with the first five natural frequencies of edge elastically restrained beams versus the elastic restraint stiffness. Then a backpropagation network is created to quickly predict the elastic restraint stiffness of beams.
關聯:
中國機械工程學會第十屆學術研討會固力組論文集=Proceedings of the Tenth National Conference of the Chinese Society of Mechanical Engineers，頁679-688
