【摘要】 以多传感器构成的复杂测量系统来监测大型桥梁健康情况已经成为目前土木、仪器科学领域研究的一个热点。桥梁结构作为一个复杂的响应系统,健康监测系统各传感器测点数据皆为这一系统的响应输出,其间必然存在一定的关联。现有的关联关系主要是从传感器测点的位置来考虑,没有反映出测点之间复杂的关联性。广义上,测点之间的关联性可以分为环境关联、结构关联、位置关联、时间关联、类别关联5类,而在数据形式上表现为正相关、负相关或零相关。以此为基础,本文结合数据挖掘和信号系统的关联理论知识,建立多传感器测点之间的关联度模型,并结合重庆马桑溪长江大桥桥梁结构健康监测系统的多传感器测量系统,分析和度量测点之间的关联关系,并在连续失效数据识别中进行试验验证。更多还原

【Abstract】 The complicated acquisition system composed of multiple sensors to monitor the structural health of the large span bridges constitutes a key research issue in civil engineering and instrumentation science. Because of the complexity of bridge structures, data from the sensors of a health monitoring system are the response output and thus would be correlated. Traditional analysis of correlation between the monitoring points only considers their positions, therefore can not reflect the complexity of reality. The correlation and form of correlation of bridge checking points are described, and by integrating the correlation theory of data mining and signal system, a correlation model is established for bridge checking points. By identifying the continuous invalid data from the Chongqing Masangxi Yangtze River Bridge structural health monitoring system, simulation results verify the effectiveness of the design method and theory.更多还原