一种基于数据融合的结构损伤识别方法

A Method of Structural Damage Identification Based on Data Fusion

目的研究多传感器的一致性数据融合算法,并应用于结构损伤识别.针对一致性数据融合方法的局限性,提出一种改进的结构识别方法.方法多个传感器同时采集结构不同位置的加速度信号可以看做由多个传感器同时测每一时刻的结构加速度值,将改进一致性数据融合算法用于结构加速度响应信号的损伤识别.结果改进的数据融合方法能够充分利用所有传感器的有效信息,在部分传感器性能降低,甚至是失效的情况下,仍能对结构损伤做出正确判断.结论多传感器数据融合包含了结构不同位置的不同状态信息,能够全面地描述结构工作状态,更有可能对结构损伤做出正确判断.

The purpose of this study was to research on the consistent data fusion algorithm of multi-sensor and applied to the identification of structural damage.For overcoming the limitation of consistency data fusion method,an improved algorithm is presented.The acceleration signals acquired simultaneously by the multiple sensors on different positions can be taken as the acceleration values which are measured simultaneously by multiple sensors at each moment.The improved algorithm of consistent data fusion utilizes the acceleration response signals to identify the structural damage.Improved method of the consistent data fusion can effectively take ad-vantage of all the sensors information.In the case of the performance degradation or even failure of some sensors,it is still possible to make right judgments on the structural damage.Multi-sensor data fusion method includes the different state information on the different positions,which can fully describe the working conditions of the structure.It makes more possible to do right judgments on the structural damage.