长江流域资源与环境 >> 2006, Vol. 15 >> Issue (6): 728-728.

• 自然资源 • 上一篇    下一篇

用Markov模型预测长江水质

胡宏昌   

  • 收稿日期:2005-10-08 修回日期:2006-02-16 出版日期:2006-11-20

FORECASTING WATER QUALITY OF THE YANGTZE  RIVER USING MARKOV MODEL

HU Hongchang   

  • Received:2005-10-08 Revised:2006-02-16 Online:2006-11-20

摘要:

由于长江水质的污染程度日益严重,为了说明治理长江对长江水质进行了简单的评价,保护长江迫在眉睫,首先,根据长江流域的17个观测站近两年多的水质检测数据统计,说明了近两年多来长江的防污治理工作有一定的效果。然后,根据1995~2004年长江流域水质的数据报告,考虑各类水之间的相互转化,构造了马尔柯夫(Markov)转移矩阵,建立了马尔柯夫预测模型,通过已有的观测数据验证了该模型的正确性及有效性。运用该马尔柯夫模型预测未来10年水质的变化趋势,即Ⅰ、Ⅱ、Ⅲ、Ⅳ类水逐年减少,而Ⅴ、劣Ⅴ类水逐年增加。到2014年,长江的第Ⅰ类水只有0.4 059%,劣Ⅴ类水达到26.2 714%,不可饮用水(即第Ⅳ,Ⅴ,劣Ⅴ类水)将达到47.468%,为此应采取更加有效的治理措施,控制长江水质的恶化。最后,通过计算,得到了每年需要处理污水的最小百分比,才能杜绝劣Ⅴ类水,将第Ⅳ、Ⅴ类水控制在20%内,从而才能保证我们有足够的饮用水

关键词: 长江水质, Markov模型, 预测

Abstract:

The water quality in the Yangtze River has become worse from time to time, and thus to control the water quality and to protect the river have been a major issue for sustainable development in the region. In this paper, a model for the prediction of water quality in the river has been established, and this will provide reference values for any control measures to be taken for the protection of the water quality. First, by using the water quality data in 17 observation stations obtained within recent two years in the Yangtze River, a simple evaluation of the water quality in the river was conducted, and it is shown that the recent measures taken for the control of the water quality in the Yangtze River had some positive effect for improving the water quality. Second, the water quality data obtained from 1995 to 2004 in the Yangtze River were used, and the Markov transferred matrix is structured, with the Markov model established. The accuracy and the validity of the model were confirmed by the observed data. The water quality in the Yangtze River in the next ten years is then forecasted by the Markov model. That is, the standards I, II, III, IV water will reduce year by year, and the V, the poor standard V water will increase gradualy. To 2014, the kind I water in the Yangtze River only has 0.405 9 percent, the poor kind V water will be 26.2714 percent, the undrinkable water (namely IV,V and the poor V water) will reach 47.468 percent. Therefore, we must find a much more effective way to control deterioration of the water quality in the Yangtze River. Finally, through the computation, the smallest percentages of sewage water to be processed each year were obtained. The poor kind V water might be prohibited, and the kinds IV and the poor V water might be controlled in 20%, thus guaranteeing that we have enough drinkingwater source.

Key words: water quality of the Yangtze River, Markov model, forecast

[1] 李杨, 李海东, 施卫省, 何俊德, 胡亚文. 基于神经网络的土壤重金属预测及生态风险评价[J]. 长江流域资源与环境, 2017, 26(04): 591-597.
[2] 刘倩倩, 陈岩. 基于粗糙集和BP神经网络的流域水资源脆弱性预测研究——以淮河流域为例[J]. 长江流域资源与环境, 2016, 25(09): 1317-1327.
[3] 龚艳冰, 戴靓靓, 胡娜, 刘高峰, 张继国. 基于云推理和模糊逻辑关系模型的干旱等级预测方法研究[J]. 长江流域资源与环境, 2016, 25(08): 1273-1278.
[4] 乔伟峰, 王亚华, 方斌. 基于变更平衡表的土地利用演化信息提取与预测方法研究——以苏州市为例[J]. 长江流域资源与环境, 2015, 24(09): 1514-1520.
[5] 韩玉刚, 曹贤忠. 皖江区域城市能级与生态环境协调度的测度和发展趋势研究[J]. 长江流域资源与环境, 2015, 24(06): 909-916.
[6] 胡鸿兴, 何伟, 沈虹, 王钰. 湖北“两圈”区域发展可持续性评价与预测[J]. 长江流域资源与环境, 2010, 19(04): 351-.
[7] 庄建琦, 崔鹏. 基于BP神经网络泥石流沟发育阶段的判定[J]. 长江流域资源与环境, 2009, 18(9): 849-.
[8] 秦建成. ArcGIS支持下的样本稀疏山区空间插值模拟探讨[J]. 长江流域资源与环境, 2009, 18(5): 489-.
[9] 罗璐琴, 周敬宣, 李湘梅. 生态足迹动态预测模型构建与分析[J]. 长江流域资源与环境, 2008, 17(3): 440-440.
[10] 杨 松, 曾永年, 吴桂平, 齐庆超, 杨 凯. 改进的最优组合人口预测模型及其应用[J]. 长江流域资源与环境, 2008, 17(3): 346-346.
[11] 彭刚华,黄良英. 长江水质评价和预测模型探讨[J]. 长江流域资源与环境, 2006, 15(Sup1): 77-82.
[12] 杨丽霞,杨桂山,苑韶峰. 数学模型在人口预测中的应用——以江苏省为例[J]. 长江流域资源与环境, 2006, 15(3): 287-291.
[13] 李 娜,夏永久. 宁波城市基础设施现代化水平综合评价与预测[J]. 长江流域资源与环境, 2006, 15(2): 136-141.
[14] 刘 伟. 长江经济带区域经济差异分析[J]. 长江流域资源与环境, 2006, 15(2): 131-135.
[15] 李崇明,黄真理. 三峡水库入库污染负荷研究(Ⅱ)——蓄水后污染负荷预测[J]. 长江流域资源与环境, 2006, 15(1): 97-106.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!