摘要
污染源自动监测数据中存在大量异常数据,严重影响数据整体质量。建立科学可靠的自动监测数据诊断分析处理方法,可有效提升在线监控监管能力水平,为数据的深度应用提供支持。但目前尚缺乏对该方法的深入研究。因此,综述了数据挖掘领域主流异常数据检测方法,并总结了在电力、交通、金融、航天等领域的应用情况,指出存在的不足和发展方向,旨在为智能污染源自动监控数据异常检测提供指导,促进污染源自动监控系统发展。
It is common to ind anomalies from automatic monitoring data of pollutant sources, which deteriorates the whole quality of the monitoring data significantly. It can effectively improve the online managing capacity to establish a set o scientific and reliable approaches for detecting, analyzing and handling automatic monitoring data. Furthermore, it can provide support for the further application of automatic monitoring data. But, it still lacks sufficient research on this approach. Therefore, this paper reviews the main anomalies detecting methods in the perspective of data mining and summarizes the important applications of a-nomalies detecting in the fields o power system, transportation, finance, and aerospace. Finally, the shortage and future ersearch direction of anomalies detecting is analyzed. This survey aims to provide a useful guidance for designing intelligent anomalies de-tecting methods of online monitoring data and promote the development of automatic monitoring system of pollutant resources.
出处
《环境科学与管理》
CAS
2016年第10期114-117,共4页
Environmental Science and Management
关键词
污染源监控
异常数据
智能检测
pollutant resource monitoring
anomaly
intelligent detecting
