改进自适应加权融合的综合管廊环境温度监测

Environmental Temperature Monitoring of Pipe Gallery Based on Improved Self-Adaptive Weighted Fusion

针对地下空间环境信息数据采集中单个传感器传输数据误差大、可靠性低容易造成环境监测系统误报警、漏报警的问题,提出了一种基于数据级融合的、不需要传感器任何先验知识的改进自适应加权融合算法。利用统计过程控制(SPC)方法剔除偏离大的传感器数据,而后通过改进一致性多传感器融合算法重新定义置信距离测度,计算各个传感器之间互相的支持程度,对于剔除的偏离较大的异常数据,用支持度最高的传感器数据代替,提高参与融合传感器数据的可靠性;最后使用自适应加权融合算法对数据进行融合,并与算数加权平均法和直接使用自适应加权融合算法的计算结果进行对比分析。结果表明:该方法融合效果好,鲁棒性强,很好的消除了由于单个传感器传输故障等造成的误报警和漏报警的可能,提高了地下空间环境信息监测的精确度,具有较好的工程实用价值。

Aiming at the problems of large errors and low reliability of the data transmitted by a single sensor in the underground space environmental information data collection,it is easy to cause false alarms and missed alarms in the environmental monitoring system.An improvement based on data-level fusion without any prior knowledge of sensors is proposed.Adaptive weighted fusion algorithm.The statistical process control(SPC)method is used to eliminate sensor data with large deviations,and then the confidence distance measure is redefined by improving the consistent multi-sensor fusion algorithm,and the degree of mutual support between each sensor is calculated.For the abnormal data with large deviations,Replace the sensor data with the highest support to improve the reliability of the sensor data involved in the fusion;finally,use the adaptive weighted fusion algorithm to fuse the data,and compare and analyze the calculation results with the arithmetic weighted average method and the direct use of the adaptive weighted fusion algorithm.The results show that the method has a good fusion effect,strong robustness,and eliminates the possibility of false alarms and missed alarms caused by a single sensor transmission failure,etc.,improves the accuracy of underground space environmental information monitoring,and has good performance in engineering practical value.