基于BP神经网络的地下水强化除砷建模研究  

Modeling for Enhanced Arsenic Removal from Groundwater Based on BP Neural Network

作　　者：段春毅[1] 赵志领[2] 杨佳露[3] 

机构地区：[1]广西建设职业技术学院设备工程系,广西南宁530003 [2]华侨大学土木工程学院,福建厦门361021 [3]中国建筑西南设计研究院有限公司,四川成都610041

出　　处：《中国给水排水》2015年第1期45-48,共4页China Water & Wastewater

基　　金：国家水体污染控制与治理科技重大专项(2009ZX07424)

摘　　要：针对郑州市东周水厂铁锰复合污染条件下的含砷地下原水,经中试和生产性试验研究,在无混凝的曝气/接触氧化过滤除砷技术的基础上,建立了投药量BP神经网络模型。该模型在以出水As<10μg/L为控制条件的前提下,进行不同原水水质条件下药剂投加量的模拟计算,并作为核心模拟算法模块,应用于水厂除砷自动控制加药系统中,完善了水厂当前的自控系统。Pilot-scale and full-scale experiments were conducted to remove arsenic from raw groundwater polluted by iron and manganese in Dongzhou Waterworks in Zhengzhou. BP neural network model for chemicals dosage control was established based on the process of aeration and contact oxidation filtration without coagulation. Under the premise of As 〈 10 Ixg/L in the effluent, simulation calculation of chemicals dosage was performed by the model under different raw water conditions. The model, as core simulation algorithm module, was applied to the automatic control system for chemicals dosing for arsenic removal to improve the current control system in the waterworks.

关 键 词：地下水 曝气/接触氧化过滤 除砷 BP神经网络模型 自控系统 

分 类 号：TU991.2[建筑科学—市政工程]