基于人工蜂群算法的BP双隐含层神经网络水质模型  

Water Quality Evaluation Model Based on Artificial Bee Colony Algorithm and BP Double Hidden Layer Neural Network

作　　者：杨咪[1,2] 徐盼盼[1,2] 钱会[1,2] 侯凯[1,2] YANG Mi;XU Panpan;QIAN Hui;HOU Kai(College of Environmental Science and Engineering,Chang'an University,Xi'an,Shanxi 710054,China;Key Laboratory of Subsurface Hydrology and Ecology in Arid Areas,Ministry of Education,Xi'an,Shanxi 710054,China)

机构地区：[1]长安大学环境科学与工程学院,陕西西安710054 [2]旱区地下水文与生态效应教育部重点实验室,陕西西安710054

出　　处：《环境监测管理与技术》2018年第1期21-26,共6页The Administration and Technique of Environmental Monitoring

基　　金：国家自然科学基金"银川平原地下水对条件变化的响应机制及合理开发利用研究"资助项目(41172212)

摘　　要：采用人工蜂群算法优化BP神经网络的初始权值和阈值,同时采用双隐含层来提高网络精度,选取DO、IMn、COD、BOD5和NH3-N作为评价指标,建立一个基于人工蜂群算法的BP双隐含层神经网络模型,并应用该模型对2012年黄河水系下河沿断面的各月监测数据进行水质评价,同时与BP神经网络、模糊层次评价方法作比较。结果表明:基于人工蜂群算法的BP双隐含层神经网络在水质评价时,均方误差小,多次运行的结果始终一致,评价结果合理有效。This paper used artificial bee colony algorithm to optimize BP neural network weights and thresholds,the double hidden layer was also used to improve the precision of the network requirements,DO,IMn,COD,BOD5 and NH3-N were selected as the evaluation index,and then a water quality evaluation model was establish based on artificial bee colony algorithm and BP double hidden layer neural network.The established model was applied for the water quality evaluation in the Xiaheyan section of Yellow River in 2012.Meanwhile,the evaluation method was compared with the BP neural network and the Fuzzy hierarchy evaluation.The results showed that the water quality evaluation based on artificial bee colony algorithm and BP double hidden layer neural network got small mean square error,and the results of multiple runs kept in accordance with each other,the evaluation results were reasonable and effective.

关 键 词：BP神经网络 双隐含层 人工蜂群算法 水质评价 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] X824[自动化与计算机技术—控制科学与工程]