基于神经网络的北京市水体水华短期预报系统
|
摘要
采用算法改进型的BP神经网络,选择叶绿素含量、磷、氮磷比、电导率和水温五个参数作为模型输入,以预测1日、3日和5日后的叶绿素含量为目标,构建了北京市长河水系水华短期预报系统。该系统三个周期的预测精度分别达到了97.2%、94%、88.3%,并且具有较好的泛化能力。相比于其它智能算法,BP神经网络结构简单、方便实用,仍然具有很强的应用性。
A water-bloom short-time predicting system of Beijing CHANGHE water system is founded with ameliorated BP neural network,in which content of chlorophyll,phosphor,ratio of nitrogen and phosphor,conductance and temperature of water are chose as the inputs of model and the target is to predict the contents of chlorophyll after 1 day,3 days and 5 days.The precisions of three periods,which the system get,reached separately 97.2%,94%,88.3%,and the system has good universality.Compared with other intelligent methods,BP neural network is simple,convenient and practicality,and has good appliance.
A water-bloom short-time predicting system of Beijing CHANGHE water system is founded with ameliorated BP neural network,in which content of chlorophyll,phosphor,ratio of nitrogen and phosphor,conductance and temperature of water are chose as the inputs of model and the target is to predict the contents of chlorophyll after 1 day,3 days and 5 days.The precisions of three periods,which the system get,reached separately 97.2%,94%,88.3%,and the system has good universality.Compared with other intelligent methods,BP neural network is simple,convenient and practicality,and has good appliance.
引文
[1]杜桂森,吴玉梅,扬忠山,等.北京城区河湖水质分析[J].湖泊科学,2005,17(4):373-377.
[2]Simeonov V,Stefanov S,Tsakovski S.Environmental treatment of water quality survey data from yangtse river[J].Mikrochim Acta,2000,134(122):15-21.
[3]Loke E,Warnaars E A,Jacobsen P,et al.Artificial neural networks as a fool in urban storm damage[J].Water Science and Technology,1997,36(829):101-110.
[4]任黎,董增川,李少华.人工神经网络模型在太湖富营养化评价中的应用[J].河海大学学报:自然科学版,2004,32(2):147-150.
[5]林志贵,徐立中,黄凤辰,等.基于D-S理论的多源水质监测数据融合处理[J].计算机工程与应用,2004,40(10):3-5.
[6]Houston L,Barbour M T,Lenat D,et al.A multi-Agency comparison of aquatic macroinvertebrate-based stream bioassessment method-ologies[J].Ecological Indicators,2002,(1):279-292.
[7]时红,孙新忠,范建华,等.水质分析方法与技术[M].北京:地震出版社,2001.
[8]金为民.千岛湖富营养化评价及其模型应用研究[D].杭州:浙江大学,2002-05:14.
[9]楼文高.湖库富营养化人工神经网络评价模型[J].水产学报,2001,25(5):474-478.
[10]曾光明,卢宏玮,金相灿,等.洞庭湖水体水质状况及运用小波神经网络对营养状态的评价[J].湖南大学学报:自然科学版,2005,32(1):91-94.
[11]吴利斌,尚士友,岳海军,等.利用模糊神经网络对湖泊富营养化程度进行评价的研究[J].内蒙古农业大学学报,2004,25(4):67-70.
[12]管军,徐立中,石爱业.基于支持向量基的水质监测数据处理及状况识别与评价方法[J].计算机应用研究,2006,9:36-38.
[13]楼文高.BP神经网络模型在水环境质量综合评价应用中的一些问题[J].水产学报,2002,26(1):90-95.
[14]韩力群.人工神经网络理论、设计及应用[M].北京:化学工业出版社,2002.
[15]阎平凡.人工神经网络与模拟进化计算[M].北京:清华大学出版社,2000.
[2]Simeonov V,Stefanov S,Tsakovski S.Environmental treatment of water quality survey data from yangtse river[J].Mikrochim Acta,2000,134(122):15-21.
[3]Loke E,Warnaars E A,Jacobsen P,et al.Artificial neural networks as a fool in urban storm damage[J].Water Science and Technology,1997,36(829):101-110.
[4]任黎,董增川,李少华.人工神经网络模型在太湖富营养化评价中的应用[J].河海大学学报:自然科学版,2004,32(2):147-150.
[5]林志贵,徐立中,黄凤辰,等.基于D-S理论的多源水质监测数据融合处理[J].计算机工程与应用,2004,40(10):3-5.
[6]Houston L,Barbour M T,Lenat D,et al.A multi-Agency comparison of aquatic macroinvertebrate-based stream bioassessment method-ologies[J].Ecological Indicators,2002,(1):279-292.
[7]时红,孙新忠,范建华,等.水质分析方法与技术[M].北京:地震出版社,2001.
[8]金为民.千岛湖富营养化评价及其模型应用研究[D].杭州:浙江大学,2002-05:14.
[9]楼文高.湖库富营养化人工神经网络评价模型[J].水产学报,2001,25(5):474-478.
[10]曾光明,卢宏玮,金相灿,等.洞庭湖水体水质状况及运用小波神经网络对营养状态的评价[J].湖南大学学报:自然科学版,2005,32(1):91-94.
[11]吴利斌,尚士友,岳海军,等.利用模糊神经网络对湖泊富营养化程度进行评价的研究[J].内蒙古农业大学学报,2004,25(4):67-70.
[12]管军,徐立中,石爱业.基于支持向量基的水质监测数据处理及状况识别与评价方法[J].计算机应用研究,2006,9:36-38.
[13]楼文高.BP神经网络模型在水环境质量综合评价应用中的一些问题[J].水产学报,2002,26(1):90-95.
[14]韩力群.人工神经网络理论、设计及应用[M].北京:化学工业出版社,2002.
[15]阎平凡.人工神经网络与模拟进化计算[M].北京:清华大学出版社,2000.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心