BP人工神经网络模型在红外线法测量含沙量中的应用研究
|
摘要
为消除泥沙粒径和泥沙亮度在红外线法测量含沙量中的干扰,建立可用于红外泥沙测量的数学模型,采用BP人工神经网络方法,将红外泥沙测量传感器输出的散射光强度值、泥沙粒径因子和泥沙亮度值作为网络的输入,对应的泥沙含量值作为网络输出,通过对网络进行训练建立测量泥沙含量的人工神经网络模型。结果表明:神经网络模型对四种不同粒度泥沙样本含沙量的测量误差均小于20.0%,当含沙量在150～275 kg/m~3时,测量误差在8.0%以内。该方法可以有效消除泥沙粒径因子和泥沙亮度因子的影响,显著提高含沙量测量结果的准确性。
In order to eliminate the interference of sediment particle size and brightness in measuring sediment concentration by infrared method, a mathematical model was established.Using BP artificial neural network algorithm, the scattered light intensity, sediment particle size factor and sediment brightness value output by infrared sediment measurement sensors are taken as the input of the network, and the corresponding sediment content value is taken as output of the network. The ANN model for measuring sediment content is established by training the network.The result shows that the measurement error of the four sediment samples with different particle sizes is less than 20.0% by using the model, and when the sediment content is between 150～275 kg/m~3, the measurement error is within 8.0%.This method can effectively eliminate the influence of the particle size factor and the brightness value of the sediment and improve the accuracy of the measurement results obviously.
In order to eliminate the interference of sediment particle size and brightness in measuring sediment concentration by infrared method, a mathematical model was established.Using BP artificial neural network algorithm, the scattered light intensity, sediment particle size factor and sediment brightness value output by infrared sediment measurement sensors are taken as the input of the network, and the corresponding sediment content value is taken as output of the network. The ANN model for measuring sediment content is established by training the network.The result shows that the measurement error of the four sediment samples with different particle sizes is less than 20.0% by using the model, and when the sediment content is between 150～275 kg/m~3, the measurement error is within 8.0%.This method can effectively eliminate the influence of the particle size factor and the brightness value of the sediment and improve the accuracy of the measurement results obviously.
引文
[1] 郑粉莉.浅谈我国土壤侵蚀学科亟待加强的研究领域[J].水土保持研究,1999,6(2):26-31.
[2] 唐克丽.中国土壤侵蚀与水土保持学的特点及展望[J].水土保持研究,1999,6(2):2-7.
[3] 高佩玲,雷廷武,赵军,等.坡面侵蚀中径流含沙量测量方法研究与展望[J].泥沙研究,2004,29(5):28-33.
[4] 张明社,李小昱,雷廷武,等.用基于人工神经网络的数据融合法测量水流泥沙含量[J].农业工程学报,2002,18(4):41-43.
[5] 任裕民,安凤玲,詹秀玲.红外测沙仪的研制和应用[J].实验技术与管理,1999,16(3):26-29.
[6] 王娅娜,蔡辉,马洪蛟,等.红外实时测沙仪研制及其应用[J].海洋工程,2007,25(3):132-135.
[7] 雷廷武,张宜清,赵军,等.近红外反射高含量泥沙传感器研制[J].农业工程学报,2013,29(7):51-57.
[8] 李勇涛,陈英智,李立新,等.基于940nm普通红外光源的反射式泥沙测量传感器研究[J].水土保持应用技术,2015(6):10-12.
[9] 张芸,张鹰,王晶晶.悬浮泥沙光谱反射率与浓度、粒径的关系[C].//第十届中国河口海岸学术研讨会论文集.南京:南京大学,2007:207-212.
[10] 唐兆民,何志刚,韩玉梅.悬浮泥沙浓度的测量[J].中山大学学报(自然科学版),2003,42(S2):244-247.
[11] 黑龙江省水土保持科学研究院.后向散射式红外泥沙测量传感器:CN201620494002.5[P].2016-12-07.
[12] 李勇涛,陈英智,李立新,等.红外泥沙测量中的多种影响因素对比分析[J].水土保持应用技术,2017(5):19-21.
[13] 刘丽颖,官冬杰,杨清伟,等.基于人工神经网络的喀斯特地区水资源安全评价[J].水土保持通报,2017,37(2):207-214.
[2] 唐克丽.中国土壤侵蚀与水土保持学的特点及展望[J].水土保持研究,1999,6(2):2-7.
[3] 高佩玲,雷廷武,赵军,等.坡面侵蚀中径流含沙量测量方法研究与展望[J].泥沙研究,2004,29(5):28-33.
[4] 张明社,李小昱,雷廷武,等.用基于人工神经网络的数据融合法测量水流泥沙含量[J].农业工程学报,2002,18(4):41-43.
[5] 任裕民,安凤玲,詹秀玲.红外测沙仪的研制和应用[J].实验技术与管理,1999,16(3):26-29.
[6] 王娅娜,蔡辉,马洪蛟,等.红外实时测沙仪研制及其应用[J].海洋工程,2007,25(3):132-135.
[7] 雷廷武,张宜清,赵军,等.近红外反射高含量泥沙传感器研制[J].农业工程学报,2013,29(7):51-57.
[8] 李勇涛,陈英智,李立新,等.基于940nm普通红外光源的反射式泥沙测量传感器研究[J].水土保持应用技术,2015(6):10-12.
[9] 张芸,张鹰,王晶晶.悬浮泥沙光谱反射率与浓度、粒径的关系[C].//第十届中国河口海岸学术研讨会论文集.南京:南京大学,2007:207-212.
[10] 唐兆民,何志刚,韩玉梅.悬浮泥沙浓度的测量[J].中山大学学报(自然科学版),2003,42(S2):244-247.
[11] 黑龙江省水土保持科学研究院.后向散射式红外泥沙测量传感器:CN201620494002.5[P].2016-12-07.
[12] 李勇涛,陈英智,李立新,等.红外泥沙测量中的多种影响因素对比分析[J].水土保持应用技术,2017(5):19-21.
[13] 刘丽颖,官冬杰,杨清伟,等.基于人工神经网络的喀斯特地区水资源安全评价[J].水土保持通报,2017,37(2):207-214.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |