用神经网络综合诊断起重机筒形支柱的损伤

Using neural network technology to synthetically diagnose damafed cylindrical supporting shell of crane

应用有限元仿真技术和人工神经网络模型 ,以反映结构损伤程度的固有频率和损伤的频率下降率作为神经网络输入的特征参数 ,对门座起重机筒形支柱上出现的裂纹损伤长度和位置进行了综合诊断和预测分析。结果表明

Taking the natural frequency and the ratio of frequency drop as input characteristic parameters,crack damage in the cylindrical supporting shell of a portal crane is diagnosed and its length and position areforecasted,applying finite element emulation and artificial neural network technologies.The result shows that the diagnostic approach presented in this paper is feasible.