无线传感器网络连续监测湿地水质异常数据的检测与处理——以西溪湿地为例  

Detection and Treatment of Anomaly Data of Continuous Monitoring of Water Quality in Wetlands by Wireless Sensor Network——A Case of Xixi Wetlands

作　　者：陈平林[1,2] 蒋锦刚[1,2] 刘婷[1,2] 

机构地区：[1]杭州师范大学遥感与地球科学研究院,浙江杭州310036 [2]浙江省城市湿地与区域变化研究重点实验室,浙江杭州310036

出　　处：《湿地科学》2014年第5期618-623,共6页Wetland Science

基　　金：浙江省重点科技创新团队项目(2010R50039和2012R10039-15);国家自然科学基金项目(41101371和41101039);大学生创新创业孵化项目(2012R421061)资助

摘　　要：为了保证无线传感网水质连续观测数据的有效性,针对数据序列中存在的异常数据问题,提出了基于差分分析的数据分段处理方法。该方法在连续观测数据一阶、二阶差分平稳性特征的基础上,利用差分序列分段统计参数进行质量控制。该方法主要针对全局差分统计质量控制的缺点,根据不同时间段的数据特征,进行有效的数据异常信息检测与质量控制。选取西溪湿地水质指标叶绿素a和蓝藻的监测数据进行分析验证,结果表明,文中提出的方法能对异常数据进行有效地检测和处理。The technology of wireless sensor network is one of the most influential forms of technology in recent years. The study proposed a method of data processing in segment based on differential analysis for the problem of abnormal data in the data sequence. The method is to control quality by using difference series of partition statistical arguments based on the character of stationary of first-order and second-order of continuous observation data. The method focuses primarily on the shortcomings of global quality control of differential statistics by effectively data detection and quality control on the basis of data characteristics at different times. In the experiment, Chlorophyll a and Cyanophyta in Xixi Wetland are selected for parsing and validation, the results indicates that the method interpreted in this paper detects and processes abnormal data effectively.

关 键 词：无线传感器网络 连续观测数据 质量控制 差分 西溪湿地 

分 类 号：X832[环境科学与工程—环境工程]