【作者】 胡顺仁； 陈伟民； 章鹏； 黄晓微；

【Author】 Hu Shunren~(1,2) Chen Weimin~2 Zhang Peng~1 (1 Key Laboratory of Optoelectronic Technology and System,Chongqing University,Chongqing 400044,China;2 Electronic Information and Automation Academy,Chongqing Institute of Technology,Chongqing 40050,China)

【机构】 重庆大学光电技术系统教育部重点实验室；

【摘要】 测量系统数据的可靠性严重制约着系统的正常运行,尤其在工程实践中会造成重大危害。实时、快速判断测量数据正确与否,对于提高测量系统的可靠度有重要意义。依据力学分析,桥梁挠度数据受温度、负载等因素的影响,数据变化范围较大,传统基于固定阈值方法作为挠度数据置信区间来判断数据正确性的方法是不适用的。本文在回归理论分析基础上,建立回归模型来描述挠度数据小段曲线,以此预测未来短时间内挠度数据的置信区间。仿真实验证明回归模型与实测曲线吻合较好(2个实验的均方误差分别为0.6847和0.4854),实测值都在预测值的置信区间范围内(当输入温度值为15.61,实测值为288.89,样本大小130时,预测的挠度置信区间值为289.0488 0.4708;样本大小32时,预测的挠度置信区间值为289. 0193 0.1376)。更多还原

【Abstract】 The data precision and reliability with mapping technology have brought forward the more and more de- sire.The measuring system reliability has restricted the system function,especially in engineer practice.It is very important significance for improving the data confidence of measuring system to judge the data right or wrong quick- ly.According to mechanics theory,the method of the calculation deflection on mechanical model of cable supported beam is restricted for exterior forces,which are random and uncertain,and its scope has changed greatly.Despite the acquisition system of the bridge structural health monitoring system(BSHMS) is designed by the advanced sen- sors,the abnormity deflection occurred by the acquisition system is unavoidable.How to judge the abnormity de- flection is very important in engineer,and the traditional method,which based on fixed threshold for deflection Confidence Interval,has became unfitted.The paper bases the regress analysis theory,and built the regression model to describe the partial deflection curve,then to predict the next deflection confidence interval.Simulation re- sults verify the effectiveness of the designed measuring method and theoretical discussions.The Mean-Square Error (MSE) of two experiment are 0.684 and 0.4854,and the true deflection value are in prediction confidence interval when the input the temperature value is 15.61,the true deflection is 288.89,the stylebook data number is 130,the prediction confidence interval is 289.0488 0.4708,and the stylebook data number is 32,he prediction confidence interval is 289.01930.1376.更多还原