论文部分内容阅读
结构故障诊断主要包括结构损伤识别、结构损伤定位、结构损伤程度的标定和评价三个方面内容。而频率、振型、频率和振型相结合的指标在结构故障诊断的三个方面各有优缺点。通过航空涡轮发动机风扇叶片建模,并在其上模拟出健康和损伤状态,选取频率、振型、频率和振型相结合的三类指标,借助ANSYS仿真与BP神经网络,验证了基于模态分析的发动机风扇叶片损伤诊断方法的可行性,并从数值上指出兼顾振型和频率的指标预测效果最优,也具有一定的工程实际意义。
Structure fault diagnosis mainly includes three aspects: structural damage identification, structural damage localization, structural damage identification and evaluation. The combination of frequency, mode shape, frequency and vibration mode has its own advantages and disadvantages in three aspects of structural fault diagnosis. By modeling the blades of the aeronautic turbine engine fan and simulating the health and damage conditions on the fan blades, three types of indicators including frequency, mode shape, frequency and vibration mode are selected. Based on the ANSYS simulation and BP neural network, The feasibility of the method for diagnosing engine fan blade damage is analyzed. And it is pointed out that the forecasting result of both mode and frequency is optimal, and it has certain practical significance.