改进型BP神经网络对民勤绿洲地下水位的模拟预测

摘要/Abstract

摘要： 以具有代表性的民勤绿洲为研究对象,以Matlab7.0为工作平台,对沙漠绿洲地下水埋深预测的三层前馈神经网络（BP神经网络）进行了改进。输入端因子选取民勤绿洲逐月灌溉量、红崖山水库下泄水量、月降水量、月蒸发量（20 cm）、月平均气温、时间序列6项,输出因子为民勤绿洲地下水位。通过在模型的输入层增加时间序列引导因子的方法使BP神经网络对输入端数据具备时间敏感性;通过Levenberg-Marquardt算法使网络误差最小化,并配合Bayesian正则化使网络的误差平方和、网络权重以及阈值平方和实现最优组合,最后使用相关系数、相对误差、效率系数等指标对模型的模拟结果进行检验。结果表明,通过以上一系列改进可以有效提高模型的模拟精度,增强模型的稳定性,并使模型具有良好的“泛化性”。

关键词: 地下水位, 人工神经网络, 时间序列, L-M算法, Bayesian正则化, 民勤绿洲

Abstract: Based on the work platform of Matlab-7.0, the paper predicted the depth of groundwater in Minqin Oasis of Gansu Province with the improved three-layers Back-Propagation Neutral Network (BP Neutral Network) model. The inputting 6 factors include monthly irrigation water volume, outflow of Hongyashan reservoir, and monthly precipitation, evaporation, air temperature, and time sequence; the output factor is the groundwater level in Minqin Oasis. The BP Neural Network Model became more sensitive to the temporal evolution when the time sequence factor was input; The Levenberg-Marquardt algorithm could minimize the bias values, and the use of the Bayesian regularization could optimize the combination of squared errors, weights and the sum of the squared threshold. The modeling results were evaluated with the correlation coefficients, relative error, efficiency index, etc. The results showed that the improved model could improve model's simulation precision and stability.

Key words: groundwater level, artificial neural network, time sequence, Levenberg-Marquardt algorithm, the Bayesian regularization, Minqin Oasis

中图分类号:

P641.7

郭瑞;冯起;翟禄新;司建华;常宗强;苏永红;席海洋. 改进型BP神经网络对民勤绿洲地下水位的模拟预测[J]. 中国沙漠, 2010, 30(3): 737-741.

GUO Rui;FENG Qi;ZHAI Lu-xin;SI Jian-hua;CHANG Zong-qiang;SU Yong-hong;XI Hai-yang. Simulation and Prediction of Groundwater Level with Improved BP Neural Network Model in Minqin Oasis[J]. JOURNAL OF DESERT RESEARCH, 2010, 30(3): 737-741.

参考文献

[1]杨自辉,俄有浩,方峨天,等.民勤绿洲边缘物种多样性对水资源变化的响应[J].中国沙漠,2007,27(2):279-283.
[2]杜建会,严平,丁连刚,等.民勤绿洲不同演化阶段白刺灌丛沙堆表面土壤理化性质研究[J].中国沙漠,2009,29(2):248-253.
[3]马国军,刘君娣,林栋,等.石羊河流域水资源利用现状及生态环境效应[J].中国沙漠,2008,28(3):592-597.
[4]汪杰,王耀琳,李昌龙,等.民勤绿洲水资源利用中的问题与节水途径[J].中国沙漠,2006,26(1):103-107.
[5]丁宏伟,王贵玲,黄晓辉.红崖山水库径流量减少与民勤绿洲水资源危机分析[J].中国沙漠,2003,23(1):84-89.
[6]马兴旺,李保国,吴春荣,等.绿洲区土地利用对地下水影响的数值模拟分析[J].资源科学,2002,24(2):49-55.
[7]孙雪涛.民勤绿洲水资源利用分析[J].中国水利,2003,12(A刊):35-38.
[8]马金珠,魏红.民勤地下水资源开发引起的生态与环境问题[J].干早区研究,2003,20(4): 261-265.
[9]徐建华.现代地理学中的数学方法[M].北京:高等教育出版社,2002.
[10]Dawson C W,Wilby R.An artificial neural network approach to rainfall-runoff modeling[J]. Hydrological sciences journal,1998,43(1):47-66.
[11]Yutaka Fukuokaa, Hideo Matsukib, Haruyuki Minamitani,et al.A modified back-propagation method to avoid false local minima[J].Neural Networks,1998,11(6):1059-1072.
[12]宁宝英,何元庆,和献中,等.黑河流域水资源研究进展[J].中国沙漠,2008,28(6):1180-1185.
[13]赵延涛,姜宝良.基于BP神经网络的地下水水位预测[J].勘察科学技术,2001(4):7-10.
[14]卢王宗志,金菊良,郑子升,等.预测济南地下水位的BP神经网络模型及其改进[J].水利水运工程学报,2005(4):71-74.
[15]祝树金,赖明勇.基于贝叶斯正则化的TDBPNN模型在中国外膜预报中的应用及评估[J].中国管理科学,2005,13(1):1-8.
[16]文喜,杨忠平,李平,等.基于改进BP算法的地下水动态预测模型[J].水资源保护,2007,23(3):5-9.
[17]苑希民,李鸿雁,刘树坤,等.神经网络和遗传算法在水可以领域的应用[M].北京:中国水利水电出版社,2002,77-80.
[18]蔡煜东,姚林声.径流长期预报的人工神经网络方法[J].水科学进展,1995,6(1):61-65.
[19]常亮,解建仓.应用优化神经网络算法预报地下水位[J].水利水运工程学报,2005,4:66-70.
[20]Hagan M T,Menhaj M.Training feed forward network with the Marquardt algorithm[J].IEEE Transactions on Neural Networks,1994,5(6):989-993.
[21]Lisboa P G J.现代神经网络应用[M].邢春颖译.北京:电子工业出版社,1996:1-85.