水资源污染负荷强度的灰色有效性预测模型及其应用
|
摘要
水资源污染负荷强度预测是水污染防治的关键环节。基于灰色系统理论,构建了水资源污染负荷强度的GM(1,1)模型、Verhulst模型和SCGM(1,1)c模型,并利用预测有效度计算各单预测模型的权重,进而建立水资源污染负荷强度的灰色GM-Verhulst-SCGM组合预测模型,在此基础上,选取2004-2013年期间工业单位产值化学需氧量排放量历史数据进行模型拟合,利用其2014-2016年数据进行模型检验。研究发现,灰色组合预测模型呈现出更低的预测误差,符合水资源污染负荷强度高精度预测需求;而通过对水资源污染负荷强度实证预测发现,其负荷强度整体上呈逐步削弱的态势,但可能会出现其高速下降向稳步趋缓转变的速率"拐点",预示着水污染防治将由"浅水区"向"深水区"的转变。
The prediction of water pollution load intensity is the key link of water pollution control. Hence, the GM(1,1) model, Verhulst model and SCGM(1,1)c model of water pollution load intensity were constructed based on grey system theory, and then the weight of these predictive models was calculated by predicting effectiveness method, and the GM-Verhulst-SCGM grey combination prediction model could be constructed. Moreover, the historical data of chemical oxygen demand(COD) emissions per unit of industrial output value was applied to model fitting from 2004 to 2013, and the data from 2014 to 2016 was applied to model test. The results showed that: the prediction error of grey combination model was lower, so it met the need of accurate prediction of water pollution load intensity. Meanwhile, empirical analysis found that the water pollution load intensity was gradually declined, but there might be an "inflection point" could emerge. It meant that the rate of water pollution load intensity descent will gradually slow down form high speed, and the water pollution control would be changed from "shallow water area" to "deep water area".
The prediction of water pollution load intensity is the key link of water pollution control. Hence, the GM(1,1) model, Verhulst model and SCGM(1,1)c model of water pollution load intensity were constructed based on grey system theory, and then the weight of these predictive models was calculated by predicting effectiveness method, and the GM-Verhulst-SCGM grey combination prediction model could be constructed. Moreover, the historical data of chemical oxygen demand(COD) emissions per unit of industrial output value was applied to model fitting from 2004 to 2013, and the data from 2014 to 2016 was applied to model test. The results showed that: the prediction error of grey combination model was lower, so it met the need of accurate prediction of water pollution load intensity. Meanwhile, empirical analysis found that the water pollution load intensity was gradually declined, but there might be an "inflection point" could emerge. It meant that the rate of water pollution load intensity descent will gradually slow down form high speed, and the water pollution control would be changed from "shallow water area" to "deep water area".
引文
[1] Sehgal V, Tiwari M K, Chatterjee C. Wavelet bootstrap multiple linear regression based hybrid modeling for daily river discharge forecasting[J]. Water Resources Management, 2014,28(10):2 793-2 811.
[2] Mouatadid S, Adamowski J. Using extreme learning machines for short-term urban water demand forecasting[J]. Urban Water Journal, 2016,14(6):630-638.
[3] Djerbouai S, Souag-Gamane D. Drought forecasting using neural networks, wavelet neural networks, and stochastic models: case of the Algerois Basin in North Algeria[J]. Water Resources Management, 2016,30(7):2445-2 464.
[4] 张岩祥,肖长来,刘泓志,等.模糊综合评价法和层次分析法在白城市水质评价中的应用[J].节水灌溉, 2015(3):31-34.
[5] 张峰,薛惠锋,宋晓娜,等.基于复合系统仿真的工业用水关联要素调控效应检验[J].干旱区资源与环境, 2018,(7):43-48.
[6] 周志青,邹国防,王磊.基于ARIMA/RBF-NN的时间序列水质预测模型研究[J].科技通报, 2017,33(9):236-240.
[7] 张峰,殷秀清,董会忠.组合灰色预测模型应用于山东省碳排放预测[J].环境工程, 2015,33(2):147-152.
[8] 刘思峰,杨英杰,吴利丰.灰色系统理论及其应用[M].北京:科学出版社, 2014.
[9] Dubkov A A, Spagnolo B. Verhulst model with Lévy white noise excitation[J]. The European Physical Journal B, 2008,65(3):361-367.
[10] Kapitza C, Lodwig V, Obermaier K, et al. Continuous glucose monitoring: reliable measurements for up to 4 days with the SCGM1 system[J]. Diabetes technology & therapeutics, 2003,5(4):609-614.
[2] Mouatadid S, Adamowski J. Using extreme learning machines for short-term urban water demand forecasting[J]. Urban Water Journal, 2016,14(6):630-638.
[3] Djerbouai S, Souag-Gamane D. Drought forecasting using neural networks, wavelet neural networks, and stochastic models: case of the Algerois Basin in North Algeria[J]. Water Resources Management, 2016,30(7):2445-2 464.
[4] 张岩祥,肖长来,刘泓志,等.模糊综合评价法和层次分析法在白城市水质评价中的应用[J].节水灌溉, 2015(3):31-34.
[5] 张峰,薛惠锋,宋晓娜,等.基于复合系统仿真的工业用水关联要素调控效应检验[J].干旱区资源与环境, 2018,(7):43-48.
[6] 周志青,邹国防,王磊.基于ARIMA/RBF-NN的时间序列水质预测模型研究[J].科技通报, 2017,33(9):236-240.
[7] 张峰,殷秀清,董会忠.组合灰色预测模型应用于山东省碳排放预测[J].环境工程, 2015,33(2):147-152.
[8] 刘思峰,杨英杰,吴利丰.灰色系统理论及其应用[M].北京:科学出版社, 2014.
[9] Dubkov A A, Spagnolo B. Verhulst model with Lévy white noise excitation[J]. The European Physical Journal B, 2008,65(3):361-367.
[10] Kapitza C, Lodwig V, Obermaier K, et al. Continuous glucose monitoring: reliable measurements for up to 4 days with the SCGM1 system[J]. Diabetes technology & therapeutics, 2003,5(4):609-614.