【作者】 李辉；

【导师】 金菊良；

【作者基本信息】 合肥工业大学 ， 水利水电工程， 2018， 硕士

【摘要】 近年来,由于城市人口的急剧膨胀、经济社会的飞速发展,很多地区对水资源的开发利用已接近甚至超过了当地水资源的承载上限,水资源短缺问题日益突出并制约着区域社会、经济的可持续发展。水资源承载力诊断评价可为水资源管理过程中识别、调控水资源超载因素提供依据,对促进人水和谐、实现区域可持续发展具有重要的现实意义。论文在分析归纳水资源承载力诊断评价概念的基础上,构建了基于联系数的区域水资源承载力诊断评价模型,并利用减法集对势诊断水资源承载力的脆弱性指标,应用于安徽省水资源承载力的动态诊断评价和空间差异诊断分析的研究中;针对水资源承载力诊断预测问题,建立了基于联系数的区域水资源承载力诊断指标组合预测模型。论文取得了如下主要结论:(1)采用“评价-诊断-调控-再评价”的研究方法论,构建了基于联系数的区域水资源承载力诊断评价模型,并对安徽省水资源承载力进行了动态诊断评价研究。结果表明:安徽省2005年-2015年的水资源承载力综合评价等级值均处于2级临界超载附近,虽存在缓慢改善的趋势、但承载状况不佳;将诊断识别出引起水资源承载状况不安全的主要指标进行调控后,再次对水资源承载力的总体安全状况进行综合评价,调控后再评价的承载状况结果得到明显改善。(2)进行了安徽省水资源承载力空间差异的诊断分析评价。结果表明:安徽省江南地区水资源承载力最好,江淮地区次之,淮北地区的水资源承载力最差;从子系统层面分析,江南地区水资源承载支撑力子系统最好,江淮地区次之,淮北地区最差,但淮北地区水资源承载压力最小;从指标层面诊断分析,人均水资源量是导致安徽省各市水资源承载状况存在空间差异的最主要因素。(3)在区域水资源承载力诊断指标预测的BP神经网络、多元线性回归模型和支持向量机模型建立的基础上,构建了基于联系数的区域水资源承载力诊断指标组合预测模型(CNCF)。结果说明:CNCF综合了BP神经网络、多元线性回归和支持向量机这3种预测方法的优势,克服了在承载力诊断指标预测中BP神经网络外延性差、多元线性回归模型预测精度低、支持向量机拟合效果差的缺点,预测精度较单一的预测模型有显著提高。综上所述,论文提出的基于联系数的区域水资源承载力诊断评价方法、组合预测方法及研究思路是合理、可行的,具有很好的普适性,可为集区域水资源承载力评价、诊断、预测为一体的综合研究提供有效的方法,为区域水资源合理配置与调控提供技术支持,对于建立水资源承载力监测预警机制具有重要意义。更多还原

【Abstract】 In recent years,due to the rapid expansion of city population,rapid economic and social development in many areas,the development and utilization of water resources in many areas is close to or even more than the upper limit of the local water resources,the shortage of water resources is becoming more and more prominent and restricts the sustainable development of regional society and economy.The diagnosis and evaluation of water resources carrying capacity can provide the basis for identifying and regulating water resources overloading factors in the process of water resource management,and it has important practical significance to promote human water harmony and realize regional sustainable development.The paper is based on the analysis and induction of the concept of water resources carrying capacity diagnosis and evaluation,we build a diagnosis and evaluation model of regional water resources carrying capacity based on the number of connections,and the set pair subtractive situation was used to identify vulnerability indexes.The model is applied to the study of the dynamic diagnostic evaluation and the spatial differential diagnosis analysis of the water resources carrying capacity of Anhui province.In order to predict diagnosis of the regional water resources carrying capacity,we set up a combination prediction model for the diagnosis index of regional water resources carrying capacity based on connection number.The following main achievements have been achieved:(1)We set a diagnosis and evaluation model of regional water resources carrying capacity based on the connection number,and study the dynamic diagnosis and evaluation of the water resources carrying capacity of Anhui province.For this study we have a new research idea,first,we evaluated the overall status of the water resources carrying capacity.Second,we identified vulnerability indexes based on the evaluation results.And then we regulated the vulnerability indexes to make them invulnerability.At last,we reevaluated of water resources carrying capacity.The results shows that the comprehensive evaluation grade of water resources carrying capacity in Anhui Province from 2005 to 2015 is near the grade 2,and it has been developed to a certain extent,but there is a tendency to improve slowly.After the vulnerability indexes are regulated,the evaluation results are significantly better than before.(2)We use the model to diagnose the spatial difference of the water resources carrying capacity of Anhui province.The results shows that the water resources in Jiangnan area of Anhui province are the best,followed by the Jianghuai area,and the water resources in Huaibei area are the worst.For subsystems,the water resource carrying capacity support subsystem of Jiangnan area is the best,followed by the Jianghuai area,and the Huaibei area is the worst but it`s water resources carrying capacity pressure subsystem is the least.For indexes,per capita water resources is the main index leading to the spatial difference of water resources carrying capacity in Anhui Province.(3)Based on the establishment of BP neural network,multivariate linear regression and support vector machine model(SVM)for predicting the diagnostic index of regional water resources carrying capacity,we construct a combination prediction model of regional water resources carrying capacity based on the number of connections(CNCF).The result shows that the SPA-CF model combines the advantages of the 3 forecasting methods such as BP neural network,multiple linear regression and support vector machine.In the prediction of diagnostic index of water resources carrying capacity,CNCF overcomes the shortcomings of BP neural network,such as poor extension,multiple linear regression model,low prediction accuracy and poor support vector machine fitting effect.The CNCF model’s prediction accuracy is more greatly improved than a single model.In summary,the proposed method based on connection number for regional water resources carrying capacity diagnosis and evaluation,combined forecasting method and research ideas is reasonable and reliable,and has universality.The method can provide an effective method for the comprehensive study of the evaluation,diagnosis and prediction of regional water resources carrying capacity and can provide technical support for the rational allocation and control of regional water resources.And it is also of great significance to the establishment of water resources carrying capacity monitoring and early warning mechanism.更多还原