基于频率和模态的变电构架损伤识别研究

Research on Damage Detection Method for Substation Gantry Structures Based on Natural Frequencies and Mode Shapes

摘要: 变电构架是变电站配电装置场地的重要构筑物，其安全可靠性与变电站密切相关。文章基于频率和模态推导了变电构架结构损失识别理论模型，且通过部分频率和模态片段来进行损伤识别，提高了灵敏度；利用传感器的先验信息，发展了一种测量点优化选择算法，以保证损伤参数识别过程的适定性。用交替进行测量点优化选择、参数识别和误差更新操作来最大限度地减少先验信息不准确所带来的识别误差对损伤识别结果的影响，尽量提高损伤识别的精度。

Abstract: Gantry Structures are very important devices whose' safety is closely related to substation. A theoretical model for detecting gantry structures' damage based on natural frequencies and mode shapes has been proposed in the present paper. In order to improve sensitivity, partial frequencies and mode shapes have been chosen to detect damage, also a measurement point selection algorithm is developed by using the prior information of sensors to solve ill-posed problem in damage identification. The optimal measurement selection algorithm, parameter identification algorithm and error updating algorithm are alternately conducted to estimate the uncertainty of the prior information of sensors, which are used to improve to the precision of identification result.