采用人工神经网络求解薄板的自由振动问题

Free Vibration Frequency of Plate Solved by Artifical Neural Network

本文提出利用神经网络的非线性表示能力求解薄板的自由振动问题，分析了标准的ＢＰ算法及多种改进措施，采用动量法和自适应调整学习率的ＢＰ算法求解，比较几种计算方法的计算结果，表明采用人工神经网络计算薄板自由振动问题方法可行。

In this paper,the nonlinear mapping of Artifical Neural Network(ANN)was adopted to solve the free vibration frequency of plate.Standard BP algorithm and various measures of improving BP algorithm were analyzed,and at last,methods of momentum and adaptive learning rate were adopted.Caculating results of different methods have good agreement,it implied that it's feasible to adopt ANN to solve the free vibration frequency of plate and the calculating results solved by ANN is satisfied.