基于风险的城市多水源柔性配置与评价研究

The Uncertainties in the urban water resources allocations are considerated as the main object of the study, and the urban water supply and water demand management system are regarded as uncertain system. Firstly, putting forward quantitative methods to describe the kinds of uncertain factors in urban water supply and water demand problems, which can be helpful for making up the study by qualitative analysis now. Secondly, establishing a urban water resources allocation model with "flexibility", which is able to deal with those complicated factors systematicly, and express multielement decision maker's opinions and preferences to the full, leading to more corresponding schemes with reality. Meanwhile，establishing a urban water resources allocation model based on risk degree evaluation, in which, multi scenario interval theory and risk assessment theory are introduced to handle the effects from different factors on program results to realize flexible allocation, the best schemes can accommodate the future changing conditions, so leading to a total optimal scheme. At last, The Xi'an city in Shaanxi province will be used as a case study of the above model. Research findings from above may offer a new method based on uncertain analysis for rational urban water resources allocation, which has great importance in theory and practice.

以城市多水源配置中的不确定性问题为研究对象，将城市多水源系统视为受多种因素影响的不确定性系统，提出城市供需水系统中各类不确定性因素的定量描述方法，弥补目前定性分析研究存在的不足；建立具有"柔性"的城市多水源配置模型，系统性地处理包含多种不确定性的定量和定性因素，并能充分体现多元决策者主体的意见和偏好，使配置结果更符合实际；引入多场景区间理论和风险评估理论，建立基于风险度的城市水资源灵活配置评价模型，能够计及各种不确定性因素对规划结果的影响，以最佳的灵活配置方案来适应未来环境的变化，从而使配置方案在总体上达到最优。以陕西省西安市为例进行模型应用研究。研究成果将为城市多水源合理配置研究提供一条不确定性分析的新途径，具有重要的理论价值和现实意义。

水资源供需矛盾已成为城市发展的制约因素，进行水资源合理配置是解决水资源问题的有效途径。然而，水资源系统本身具有不确定性和复杂性，尤其是西北地区城市，由于气候干旱、降雨量少且分布不均，城市水资源短缺问题显著，水资源系统中的不确定性因素错综复杂，传统的配置方法大多难以充分反映此特征。本研究将不确定性理论和方法的应用于城市水资源系统，以更好地定量表述供需水测和配置过程中的的不确定性问题，使配置结果更符合实际，并能体现多元决策主体的意见和偏好，主要研究内容和成果如下：.（1）进行了考虑不确定性的城市需水预测方法研究，分别构建了基于盲数理论、灰色自记忆原理和径向基函数（RBF）神经网络的城市需水量预测模型。.（2）建立了两阶段随机规划水资源配置模型，第一阶段决策是对规划年不确定事件预先判断，确定各供水源的供水目标。第二阶段决策通过对未来不确定性的处理，预先决策给予补偿，使得经济损失最小。.（3）将两阶段随机规划与整数规划、区间数方法和模糊集等相结合分别建立了基于区间两阶段随机整数规划（ITSIP）和区间模糊两阶段随机规划（IFTSP）的城市多水源多用户的配置模型。.（4）将区间数和多阶段随机规划相结合建立区间多阶段随机规划模型（IMSLP），该模型能反映系统的不确定性，并通过离散情境反映水资源优化配置决策的动态特征。.（5）以陕西省城市西安市和宝鸡市为例，进行了模型应用和配置结果对比评价分析。