Nowadays,the online monitoring system of key pollution sources has been constructed by many city.In the meantime,water quality automatic monitoring system was started to use in monitoring major sources of drinking water,which obtained the real-time data of surface water quality.Using data mining technology,these data of surface water quality has a very important significance on analysis and forecast of water quality for the city.In this paper,we mainly use the automatic monitoring of data of Xiaolan waterway,Madafen monitoring station and the key pollution sources upstream and downstream as the research object,conducting data mining technology,to carry out screen,pretreatment and attribute reduction.Then the BP neural network technology is used on model building to forecast CODMn,which is the key index of water quality of Xiaolan waterway.The Analysis of forecasting results show that the error is less and the forecasting effect is better.%当前,许多地方已经建设重点污染源在线监控系统,收集了重点污染源大量的排污数据。同时,也开始建设地表水质自动监测系统,并获得地表水质的实时数据。利用数据挖掘技术,这些数据对城市地表水质的分析和预测有着十分重要的意义。本文主要以小榄水道马大丰水厂水质自动监测子站的历史数据和周边上下游重点污染源在线监控数据为研究对象,并利用数据挖掘技术,对监测数据进行筛选,预处理和属性约简,再运用BP神经网络技术建立模型,对小榄水道的主要指标CODMn进行预测。预测结果分析表明,误差较少,预测效果较好。
展开▼
