机械结构的损伤发展到一定程度,会对其功能产生严重影响,所以及早检测出损伤非常必要。针对机械结构在损伤发生时其内部能量分布会发生变化的情况,对加速度计测得的结构振动信号进行短时傅立叶变换,获得可以反映能量分布的时频谱图。使用脉冲耦合神经网络提取时频谱图的熵序列特征,很好地识别了结构的损伤。使用上述方法对简支梁的损伤进行了实验研究,验证了其有效性和实用性。
The damage in mechanical structure,if develeped to some extent,would have serious impact on the function of the mechanical structure. There fore, early detection of damage is necessary. When structure is damaged, the internal energy distribution is often changed. The time-spectrum,which is obtained by short-time Fourier transform of the vibration signals, could reflect the energy distribution of structure. In this paper, by using the method of short-time fourier transform and pulse coupled neural network, the characteristics of the structural vibration signal are extracted to identify structural damage. Using the method, the damage of simple beam was experimentally studied, and the effectiveness and practicality were verified.