多传感器自动站的数据融合效果及优势分析  

Data Fusion Effect and Advantage Analysis of Multi-Sensor Automatic Weather Station

作　　者：何艳丽[1] 黄飞龙[1] 

机构地区：[1]广东省大气探测技术中心,广州510080

出　　处：《气象》2015年第8期1028-1035,共8页Meteorological Monthly

基　　金：广东省气象局科学技术研究项目(2012B19)资助

摘　　要：采用三个同类型的温度、湿度和雨量传感器分别对气温、相对湿度以及降雨量进行测量。通过计算同一气象要素三个测量值之间的离散度以及数据集中度,结合观测数据的时间序列变化趋势,融合得到一个更具有代表性,更加准确的数据作为该气象要素的观测值。对三种气象要素多种天气状况下的融合数据进行业务对比,一致率计算结果证明,多传感器融合结果与业务观测数据一致性比较高。对差值的规律和异常数据进行分析,显示多传感器具有以下优势:在单一传感器故障时保障观测数据连续不间断,剔除由信号干扰或者误操作导致的错误数据,增加强对流过程中奇异数据的可信度,及时反映单个传感器性能变化。因此多传感器自动站不但可以作为业务使用,而且对解决业务观测中部分数据不可信的问题具有重要意义。Three sensors of the same type are adopted to monitor air temperature,relative humidity and rainfall.Then dispersion and concentration of observation data are calculated.Combined with the consideration of variation tendency in time series,models of data fusion are invented,and typical and accurate data are worked out as final meteorological observation data.The fusion data of three meteorological elements are collected in different weather conditions,and by the comparison of the data,the calculation result of concordance rate proves that there is high consistency between the two sets of data.Analysis of regularity and abnormal data of difference values show multi-sensor automatic weather station（AWS） has advantages as follows.It can make sure that observation data are continuous when single sensor is out of work;it can eliminate the influence of wrong data from interference signal and misoperation;and it can verify the validity of singular data in severe convective weather processes and reflect the performance changes of single sensor.So,the multi-sensor AWS can be applied in surface weather observation,which is of great importance for solving the observation data problems at present.

关 键 词：多传感器 自动气象站 数据融合 离散度 数据集中度 

分 类 号：P414[天文地球—大气科学及气象学] P412