为快速、准确定位供水管网异常漏损所在位置,减少水资源流失和降低漏损检测成本,以 A 村为例,在区域管网分区计量的基础上,采用遗传算法(GA)优化支持向量回归(SVR)模型,建立基于 GA-SVR 的水量预测模型,进而分析模型预测水量与实际水量之间的差异性,从而识别区域管网异常漏损情况,构建区域管网异常漏损检测模型.结果显示,基于 GA-SVR的水量预测模型,其测试期的平均纳什效率系数为 0.891;管网异常漏损识别准确率为 91.7%.结果表明,构建的 GA-SVR管网异常漏损检测模型,其异常漏损识别程度较高,实际应用效果良好,结合区域管网分区计量方法,可实现漏损的快速识别和定位.
Abnormal Leakage Detection of Pipe Network Based on GA-SVR
In order to quickly and accurately locate the location of abnormal leakage of water supply pipe network,re-duce loss of water resources and pipe network maintenance cost,taking A village for an example,on the basis of regional pipe network zoning metering,a genetic algorithm(GA)was used to optimize a support vector regression(SVR)model.And then a water volume prediction model based on GA-SVR was established to analyze the difference between predicted and actual water volumes.The abnormal leakage in the regional network was identified and a model for detecting abnor-mal leaks in the regional network was established.The results show that the average Nash efficiency coefficient(NSE)of the water quantity prediction model based on GA-SVR is 0.891.The identification accuracy of abnormal leakage of pipe network is 91.7%.The proposed GA-SVR pipe network abnormal leakage detection model has a high degree of abnormal leakage identification and good practical application effect.Combined with the regional pipe network zoning metering method,it can achieve rapid identification and positioning of leakage.
万方数据
维普
