摘要
理论分析表明,结构损伤前后固有频率的变化包含了结构损伤位置和程度的信息。在此理论基础上,对一个一端固定轴模型进行了损伤数值模拟,提取固有频率的变化信息并采取合适的方法构造改进型BP神经网络来判断结构损伤。检测表明,该方法在结构损伤检测中具有较好的应用前景。
It has been proved that the natural frequency changes of structures contain the information such as location and degree of structural damage. Based on this theory , a experiment to simulate the damage of a shaft with one end fixed is used , extracting the natural frequency changes and forming the input vectors of the BP algorithm network , detecting the structural damage by the t rained networks , the result s show the effectiveness of this method.
出处
《沈阳航空工业学院学报》
2009年第1期42-45,30,共5页
Journal of Shenyang Institute of Aeronautical Engineering
关键词
损伤识别
固有频率
改进型BP神经网络
一端固定轴
damage identification
natural frequency
BP artificial network
shaft with one end fixed
