供水管网计量区水量数据的特征分析

Characteristic analysis of water consumption data in metering zones of water distribution systems

基于供水管网计量区监测数据快速分析与挖掘,进行近实时异常工况检测,从时间序列角度解析了供水管网计量区水量的数据特征。结果表明,水量数据的波动变化可归结为规律性波动变化、不确定性波动变化以及错误数据三类,为数据驱动的异常工况检测方法开发提供技术支撑。

Monitoring data reflect the operation state of water distribution systems.With the development of information and communication technology,monitoring data can be transmitted in near real time.Under this condition,it becomes possible to detect abnormal conditions in near real time through the rapid analysis and mining of the monitoring data in metering zones.This paper uses statistical analysis method to analyze the data characteristics of water consumption data in metering zones from the perspective of time series.The results show that the fluctuation of water consumption data can be divided into three types:regular fluctuation,uncertainty fluctuation and error data.This lays a foundation for developing data-driven anomaly detection methods.