基于GIS与RBF神经网络的水厂原水藻类预测  

Forecast of Algae Content in Raw Water of Waterworks Based on GIS and RBF Neural Network

作　　者：刘俊萍[1] 马晓雁[1] 

机构地区：[1]浙江工业大学建筑工程学院,浙江杭州310032

出　　处：《中国给水排水》2015年第9期66-69,共4页China Water & Wastewater

基　　金：浙江省自然科学基金资助项目(Q12E080063;LY14E090007);国家自然科学基金资助项目(51208468);浙江工业大学重中之重学科开放研究基金资助项目

摘　　要：利用地理信息系统(GIS)对水厂原水水质信息进行管理,将水质信息与水厂的空间信息有机联系,可提高水厂水质管理水平。采用径向基函数(RBF)神经网络对杭州市的4座水厂原水中的藻类含量进行预测,建立藻类含量预测模型,结果表明,RBF神经网络藻类含量预测模型与常用的细胞计数法相结合,可提高藻类含量预测精度,同时采用GIS技术将藻类含量预测结果以空间图形形式显示和输出,更具可视性,可为水厂有效地控藻除藻提供支持。Geographic information system （GIS） was applied to manage information of raw water quality of waterworks, and associating the information of raw water with spatial information of waterworks could raise the management level of water quality. Radial basis function （RBF） neural network was used to forecast the algae contents in raw water of four waterworks in Hangzhou City, and algae content forecast model was established. The forecast results showed that the algae content forecast model based on RBF neural network combined with the commonly used cytometry could improve the forecast accuracy of algae content. Applying GIS to display and output the forecast result of algae content in the form of spatial graph was more visual, It could provide support for algae control and removal in waterworks.

关 键 词：水厂原水 藻类 地理信息系统 径向基函数神经网络 

分 类 号：TU991.2[建筑科学—市政工程]