
基于BP神经网络替代模型的地下水污染随机模拟
葛渊博, 卢文喜, 王梓博, 王涵, 常振波
基于BP神经网络替代模型的地下水污染随机模拟
Random Simulation of Groundwater Pollution Based on BP Neural Network Substitution Model
为分析水文地质参数的不确定性对地下水数值模拟模型输出结果的影响,针对假想算例进行研究。首先建立地下水数值模拟模型。运用灵敏度分析法筛选出对模拟模型输出结果影响较大的参数作为随机变量,为减少反复调用模拟模型产生的计算负荷,分别采用克里格法和BP神经网络方法建立模拟模型的替代模型并比较二者的精度,并选择精度较高的BP神经网络替代模型进行蒙特卡罗模拟。其中,采用BP神经网络方法建立替代模型时,使用三分法算法快速地确定了使替代模型误差最小的隐含层节点数。最后,对随机模拟的结果进行统计分析与区间估计,评价了地下水污染风险。结果表明:在置信水平为90%时,三口观测井浓度的置信区间分别为367.48~415.67、205.12~230.33、118.85~132.82 mg/L。结合风险评估,计算出研究区内一号/二号/三号观测井地下水遭受污染的风险分别为0.66、0.60、0.58,籍此为地下水污染防治提供科学依据。
In order to analyze the influence of the uncertainty of hydrogeological parameters on the output of the groundwater numerical simulation model, this paper studies a hypothetical example. First, a groundwater numerical simulation model is established. Then sensitivity analysis is used to screen out the parameters that have a greater impact on the output of the simulation model. As a random variable, in order to reduce the computational load caused by repeatedly calling the simulation model, the Kriging method and the BP neural network method are used to establish the replacement model of the simulation model and the accuracy levels of the two are compared, and the BP neural network replacement model with higher accuracy is selected to perform Monte Carlo simulation. Among them, when the BP neural network method is used to establish the replacement model, the third-point algorithm is used to quickly determine the number of hidden layer nodes that minimize the error of the replacement model. Finally, the results of the random simulation are statistically analyzed and the interval estimation and evaluation of the risk of groundwater pollution are compared. The results show that when the confidence level is 90%, the confidence intervals for the concentration of the three observation wells are 367.48~415.67, 205.12~230.33 and 118.85~132.82 mg/L. Combined with the risk assessment, it is calculated that the risk of groundwater pollution in observation wells No.1, No.2 and No.3 in the study area are 0.66, 0.60, and 0.58, which provides a scientific basis for groundwater pollution prevention and control.
地下水污染随机模拟 / 不确定性分析 / 替代模型 / 隐含层节点数 / 风险评估 {{custom_keyword}} /
stochastic simulation of groundwater pollution / uncertainty analysis / alternative model / number of hidden layer nodes / risk assessment {{custom_keyword}} /
表1 研究区水文地质参数数值Tab.1 Values of hydrogeological parameters in the study area |
渗透系数/(m·d-1) | 纵向弥散度/m | 孔隙度 | 给水度 | |
---|---|---|---|---|
北部地区 | 30 | 40 | 0.3 | 0.25 |
南部地区 | 35 | 50 | 0.3 | 0.3 |
表2 灵敏度近似计算解Tab.2 Sensitivity approximate calculation solution |
参数 | -0.2 | -0.1 | 0 | 0.1 | 0.2 |
---|---|---|---|---|---|
K 1 | 1.684 8 | 1.530 7 | 0 | 1.294 2 | 1.202 0 |
K 2 | 1.760 1 | 1.594 7 | 0 | 1.346 1 | 1.252 7 |
μ 1 | 0 | 0 | 0 | 0 | 0 |
μ 2 | 0 | 0 | 0 | 0 | 0 |
α 1 | 1.166 1 | 1.066 7 | 0 | 0.910 1 | 0.850 7 |
α 2 | 0.337 1 | 0.323 2 | 0 | 0.286 7 | 0.270 2 |
n | 0.000 1 | 0.000 1 | 0 | 0.001 3 | 0.000 3 |
表3 参数概率分布及取值情况Tab.3 Probability distribution and value of parameters |
随机变量名 | 概率分布 | 均值 | 取值范围 |
---|---|---|---|
渗透系数K1 (m·d-1) | 对数正态分布 | 30 | (24,36) |
渗透系数K2 (m·d-1) | 对数正态分布 | 35 | (29,41) |
纵向弥散度α 1 | 正态分布 | 40 | (36,44) |
纵向弥散度α 2 | 正态分布 | 50 | (46,54) |
表4 替代模型精度分析Tab.4 Precision analysis of substitution model |
评价指标 | R 2 | MSE | MAPE |
---|---|---|---|
克里格替代模型 | 0.999 995 | 0.049 9 | 0.078 1 |
BP神经网络替代模型 | 0.999 999 | 0.000 1 | 0.002 9 |
表5 单样本柯尔莫戈洛夫-斯米诺夫检验结果表 (mg/L)Tab.5 Results of Kolmogorov-Sminov test for a single sample |
一号井污染物浓度 | 二号井污染物浓度 | 三号井污染物浓度 | ||
---|---|---|---|---|
个案数 | 1 000 | 1 000 | 1 000 | |
正态参数 | 平均值 | 391.577 | 217.725 | 125.833 |
标准偏差 | 18.679 | 9.770 | 5.415 | |
最极端差值 | 绝对 | 0.019 | 0.023 | 0.018 |
正 | 0.019 | 0.023 | 0.018 | |
负 | -0.017 | -0.014 | -0.014 | |
检验统计 | 0.019 | 0.023 | 0.018 | |
渐近显著性(双尾) | 0.200 | 0.200 | 0.200 |
表6 三口观测井污染物浓度统计Tab.6 Pollutant concentration statistics of the three observation Wells |
井编号 | 平均值 | 中位数 | 方差 | 标准偏差 | 变异系数/% |
---|---|---|---|---|---|
一 | 391.58 | 391.27 | 348.89 | 18.68 | 4.8 |
二 | 217.73 | 217.46 | 95.45 | 9.77 | 4.5 |
三 | 125.83 | 125.84 | 29.32 | 5.42 | 4.3 |
表7 三口观测井污染物浓度区间估计Tab.7 Estimation of pollutant concentration interval for three observation Wells |
井编号 | 置信水平 | |||
---|---|---|---|---|
>90% | >75% | >60% | >45% | |
一 | 367.48~415.67 | 378.88~404.28 | 386.72~396.43 | 389.34~393.82 |
二 | 205.12~230.33 | 211.08~224.37 | 215.19~220.27 | 216.55~218.90 |
三 | 118.85~132.82 | 122.15~129.51 | 124.42~127.24 | 125.18~126.48 |
1 |
翟建桥.谈水利工程建设对生态环境的影响分析[J].清洗世界,2021,37(7):98-99.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
2 |
刘士明.水利工程地下水环境影响评价[J].河南水利与南水北调,2018,47(8):52-53.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
3 |
唐国品,牛玉珍.中国城市地下水污染演变趋势研究[J].四川水利,2021,42(1):80-83.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
4 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
5 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
6 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
7 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
8 |
王涵,卢文喜,李久辉,等.地下水DNAPLs污染多相流的随机模拟及其不确定性分析[J].中国环境科学, 2018,38(7):2 572-2 579.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
9 |
常振波,卢文喜,辛欣,等.基于灵敏度分析和替代模型的地下水污染风险评价方法[J].中国环境科学,2017,37(1):167-173.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
10 |
辛欣,卢文喜,罗建南,等.DNAPLs 污染含水层多相流数值模拟模型的替代模型[J].吉林大学学报,2011,41(3):855-860.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
11 |
王梓博,卢文喜,王涵,等.某钼矿尾矿库地下水污染的随机模拟[J].中国环境科学,2020,40(5):2 124-2 131.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
12 |
胡伟,李晓,侯佳均,等.基于BP神经网络法的平谷地下水水质评价[J].甘肃水利水电技术, 2020,56(9):13-17,62.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
13 |
潘紫东,卢文喜,范越,等.基于模拟-优化方法的地下水污染源溯源辨识[J].中国环境科学, 2020,40(4):1 698-1 705.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
14 |
束龙仓,王茂枚,刘瑞国,等.地下水数值模拟中的参数灵敏度分析[J]. 河海大学学报(自然科学版), 2007,35(5):491-495.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
15 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
16 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
17 |
杨文忠,梁凡,李东昊,等.基于BP神经网络的车辆换道时间预测[J].现代电子技术, 2020,43(21):86-92.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
18 |
束龙仓,朱元生.地下水资源评价中的不确定性因素分析[J].水文地质工程地质,2000(6):6-8.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
19 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
20 |
王嵘冰,徐红艳,李波,等.BP神经网络隐含层节点数确定方法研究[J].计算机技术与发展, 2018,28(4):32-25.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
{{custom_ref.label}} |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
/
〈 |
|
〉 |