Electric power system involves different fields and disciplines which addressed the economic system, energy system, and environment system. Inner uncertainty of this compound system would be an inevitable problem. Therefore, an inexact multistage fuzzy-stochastic programming (IMFSP) was developed for regional electric power system management constrained by environmental quality. A model which concluded interval-parameter programming, multistage stochastic programming, and fuzzy probability distribution was built to reflect the uncertain information and dynamic variation in the case study, and the scenarios under different credibility degrees were considered. For all scenarios under consideration, corrective actions were allowed to be taken dynamically in accordance with the pre-regulated policies and the uncertainties in reality. The results suggest that the methodology is applicable to handle the uncertainty of regional electric power management systems and help the decision makers to establish an effective development plan.

Abstract Economic restructuring, water resources management, population planning and environmental protection are subjects to inner uncertainties of a compound system with objectives which are competitive alternatives. Optimization model and water quality model are usually used to solve problems in a certain aspect. To overcome the uncertainty and coupling in reginal planning management, an interval fuzzy program combined with water quality model for regional planning and management has been developed to obtain the absolutely “optimal” solution in this study. The model is a hybrid methodology of interval parameter programming (IPP), fuzzy programing (FP), and a general one-dimensional water quality model. The method extends on the traditional interval parameter fuzzy programming method by integrating water quality model into the optimization framework. Meanwhile, as an abstract concept, water resources carrying capacity has been transformed into specific and calculable index. Besides, unlike many of the past studies about water resource management, population as a significant factor has been considered. The results suggested that the methodology was applicable for reflecting the complexities of the regional planning and management systems within the planning period. The government policy makers could establish effective industrial structure, water resources utilization patterns and population planning, and to better understand the tradeoffs among economic, water resources, population and environmental objectives.

为解决现有污水处理技术优选方法的片面性、主观性和不确定性,本研究以环境、社会和经济效益最大化为目标,采用区间两阶段随机规划的方法,构建模型测算区域新增废水的最优处理方法和处理量,并应用于滇池盘龙江流域,结果表明在废水排放量较大的区域,多种污水处理技术的组合要优于选择任意一种处理技术.其中,厌氧-缺氧-好氧工艺由于运营和投资成本低,在处理规模较大时,具有明显优势,而间歇式循环延时曝气活性污泥法则适用于废水排放规模较小的城镇,其污染物去除效果要优于厌氧-缺氧-好氧工艺和氧化沟法.

目前湖泊营养盐浓度级别评价常采用的平均值法忽视了变量的不确定性.本文据此提出一种基于正态分布总体均值假设检验的级别评价方法,并结合我国地表水环境质量标准(GB3838—2002)对营养盐浓度的分类,提出了简便快捷的图示法.结果发现,该方法能体现营养盐浓度变量的不确定性,融合先验级别假设,将决策风险定量化,可提供较为广阔的决策空间.同时,采用该方法对滇池外海观音山监测断面氨氮浓度的水质级别进行评价,发现先验级别假设、样本容量和决策偏好对评价结果有显著影响;与平均值法的比较发现,假设检验法具有更小的错误概率(例如,样本容量>13时,风险<0.21),评价结果具有更高的可信度,验证了假设检验法的适用性.

水质达标评价是流域水污染防治决策的科学基础,但水质变量的不确定性在传统的达标评价中常被忽略并由此产生决策风险.二项分布检验法可定量表征水质变量不确定性导致的决策风险,其结果更具鲁棒性.基于二项分布检验法,纳入风险偏好和利益权衡,提出了水质达标评价的决策框架.以海河流域为研究对象,对其2014年7月22日—2015年7月14日的周监测数据进行达标评价分析,对比了不同方法的决策风险,发现平均值法会导致很大的取伪错误概率,超标比例法会导致很大的弃真错误概率;二项分布检验法的2类错误概率均可控制在相对较低水平,决策风险最小,可为决策者提供较大的决策空间.案例研究表明,二项分布检验法在水质达标评价中具有适用性和灵活性.

Abstract Economic restructuring, energy planning and environmental protection are subject to inherent uncertainties in a compound system with competing decision objectives. Therefore, an inexact multi-objective programming model for regional economy-energy-environment system management has been developed to obtain absolutely “optimal” solutions. Under two comparative scenarios, three subsystems, six industries, four types of energy, and three kinds of air pollution were considered in an optimization model, and a net system benefit and trade-off analysis between subsystems was conducted. The methods of interval-parameter programming and multi-objective programming were incorporated into the model to tackle the uncertainties and complexities reflected in the case study. The model results indicated that the developed model could provide effective linkages among the economy-energy-environment systems and offer decision makers great insight into the reliability tradeoffs for the adjustment of the existing management policy.

不确定性是模型研究中的重要内容，虽然目前对于不确定性的研究已经出现了很多方法，但是复杂模型的不确定性的系统研究却依然匮乏。《三维水动力水质模型不确定性研究》针对三维水动力水质模型提出一套“不确定性评价-参数自动估计-决策响应可能性评价”研究体系。《三维水动力水质模型不确定性研究》以环境流体动力学模型为建模平台，以滇池湖泊为研究案例，通过构建滇池三维水动力水质模型开展不确定性的研究。首先采用Morris敏感性分析等方法对模型进行全局不确定性和敏感性分析，识别模型的不确定性分布和主控因子，以及模型不确定性的时空差异分析；然后，提出基于BP神经网络替代模型的多目标参数自动率定方法，降低参数率定的计算成本；很后，提出基于不确定性的水质响应可能性评价方法，在模型应用中考虑不确定性的影响水平。《三维水动力水质模型不确定性分析》可作为数理建模、水环境模型、水文模型等模型应用和研究领域的学者、技术人员和学生以及从事水环境管理、水文预报、环境科学等领域工作人员的参考用书。