通过对给水管网实测计量小区(DMA)的夜间流量数据进行解析, 可以及时获得近似的真实漏损数值。利用大频率、高精度的夜间实测流量数据进行分析，发现夜间不同时段DMA流量近似满足正态分布,其最小值近似接近真实漏损。根据某案例凌晨2:00-4:00夜间最小流量数据，采用置信水平95.5%、置信区间 进行了分析。结果表明，利用 表征剔除异常值后的DMA最小流量，可以消除实测数据的误差和干扰影响，来得出小区的真实漏损水量。这种方法可减少对用户夜间用水估计产生的误差,也可便利地通过流量设备进行现场计量,从而快速评估DMA的夜间漏失水平

The analysis of the night minimum flow data in the Districted Metered Area (DMA) of the water distribution system can characterize the real loss of DMA. The high-frequency and high-accuracy measured night flow data were analyzed, which indicated that the night flow data of DMA at different time periods are in normal distribution and the lowest night flow data is close to the real loss of DMA. In this study, the minimum night flow data from 2 AM to 4 AM in the early morning were analyzed based on the confidence level of 95.5% and confidence interval of (m-2d, m+2d).. The results showed that the minimum night flow of DMA after removing the abnormal values characterized by m-2dcan eliminate the effects of error and interference of measured data, and therefore obtain the real loss of DMA. This method can lower the evaluation error of night water demand, and facilitate the in-situ meter flow measurement to assess the DMA leakage level quickly.

国家水体污染控制与治理科技重大专项课题“南水北调受水区饮用水安全保障技术研究与综合示范”（2012ZX07404-004）