基于BP人工神经网络的红旗泡水库富营养化评价  

Eutrophication assessment of Hongqipao Reservoir based on Back Propagation(BP)artificial neural network

作　　者：张怡 冯萱 王硕[2] 姜海波[3] 郑国臣 ZHANG Yi;FENG Xuan;WANG Shuo;JIANG Hai-bo;ZHENG Guo-chen(College of Plant Science,Jilin University,Changchun 130062,China;School of Environment and Civil Engineering,Jiangnan University,Wuxi 214122,China;School of Environment,Northeast Normal University,Changchun,130117,China;Hebei University of Environmental Engineering,Qinhuangdao 066102,China)

机构地区：[1]吉林大学植物科学学院,吉林长春130062 [2]江南大学环境与土木工程学院,江苏无锡214122 [3]东北师范大学环境学院,吉林长春130117 [4]河北环境工程学院,河北秦皇岛066102

出　　处：《环境生态学》2022年第9期103-107,共5页Environmental Ecology

基　　金：城市水资源与水环境国家重点实验室自主课题基金(2020DX06)资助。

摘　　要：红旗泡水库是大庆市的重要水源地之一,以供应石油化工用水为主,兼顾农业灌溉、渔业养殖、生活用水等。BP人工神经网络能够很好地处理模糊的非线性关系,对于具有随机、非线性特点的水质评价和预测等复杂问题具有很好的适应性,目前被广泛应用。以红旗泡水库为研究对象,根据2016—2020年水质监测数据,以总氮、总磷、叶绿素a、高锰酸盐指数和透明度为评价指标,应用BP人工神经网络模型对水库富营养化状况进行评价。结果表明,2016—2020年红旗泡水库多处于中度富营养状态,但富营养化程度逐年降低。BP人工神经网络模型可靠性强、运行稳定、方便快捷,能够快速、准确评价湖库富营养化状况。Hongqipao reservoir is one of the important water sources in Daqing city.It mainly supplies water for petrochemical industry,agricultural irrigation,fishery breeding and domestic water.BP artificial neural network can deal with the fuzzy nonlinear relationship well.It has good adaptability to complex problems such as water quality evaluation and prediction with random and nonlinear characteristics,and is widely used.This study takes Hongqipao reservoir as the research object.According to the water quality monitoring data from 2016 to 2020,taking total nitrogen,total phosphorus,chlorophyll a,permanganate index and transparency as evaluation indexes,BP artificial neural network model is applied to evaluate the eutrophication status of the reservoir.The evaluation results show that Hongqipao reservoirs are most in moderate eutrophication from 2016 to 2020,but the degree of eutrophication decreases year by year.BP artificial neural network model has strong reliability,stable operation,convenience and rapidity,and can quickly and accurately evaluate the eutrophication status of lakes and reservoirs.

关 键 词：BP人工神经网络 富营养化 水质评价 红旗泡水库 

分 类 号：X824[环境科学与工程—环境工程]