基于LOF模型的输水管网异常数据检测及校正

Detection and Correction of Abnormal Data in Water Distribution Networks Based on LOF Model

为了解决实测流量压力数据错误对输水管道实时水力模型EPANET-RTX造成的干扰,针对大量实测数据,通过局部异常因子LOF算法与多个局部测量模型,建立异常数据检测模型检测异常样本;再采用Z分数算法检出异常参数,并采用k邻域均值法校正参数,实现高维向量快速检异、校正,从而保证EPANET-RTX稳定可靠地跟踪输水管网实时运行状态。以JS输水管网为实例进行运行数据异常检测及校正,异常数据识别准确率达99.53%,校正数据的质量及检异/校正速度均满足实时水力模型所需。

The measurement errors in flow and pressure data often interfere the real-time hydraulic model EPANET-RTX for water transmission pipelines.To address this issue,the Local Outlier factor(LOF)algorithm is used to detect abnormal samples,then the Z-score algorithm is used to detect abnormal parameters,and the k-neighborhood mean method is used to correct abnormal parameters,based on a large amount of measurement data.This approach achieves fast and accurate high-dimensional vector detection and correction,thereby ensures EPANET-RTX stably and reliably tracking the real-time hydraulic status of water transmission pipelines.Taking the JS water transmission pipeline network as an example,abnormal data detection and correction are carried out.The accuracy rate of abnormal data recognition reaches 99.53%,both the quality and the speed of correction data meet the needs of real-time hydraulic models.