【作者】 叶季平；

【导师】 王丽萍；

【作者基本信息】 华北电力大学（北京） ， 管理科学与工程， 2010， 博士

【摘要】 水库是河流综合开发治理中普遍采用的有效工程措施。利用水库调蓄洪水、削减洪峰,对提高江河防洪标准,减轻或避免洪水灾害,起着十分重要的作用。近几年,我国水资源短缺与洪水灾害频繁出现,以生态环境保护为目标的水库生态调度越来越引起社会各界人士的注意,防洪调度难以满足现在各方面需求的局面日益凸显,因此,深入研究混联水库群防洪调度管理对我国水资源可持续利用,减少自然灾害,促进社会、经济、生态协调发展具有重要的意义。本文以混联水库群防洪系统为研究对象,以对洪水预报与防洪调度管理为前提,应用相关智能算法、风险分析和可拓学等方法理论,深入探讨了混联水库群防洪调度管理中的运行优化、风险分析、方案决策和生态调度等问题,并取得了如下成果：(1)改进了洪水预报技术及参数优选方法。从洪水成因、水文模型及参数优选方法入手,在马斯京根模型参数估计问题中,引进绝对残差绝对值和最小准则,直接优选流量演算系数,且结合遗传算法的原理步骤及算法本身的参数优化特性,提出了马斯京根模型参数估计自适应加速遗传算法。(2)建立了混联水库群防洪优化调度模型与方法。论文梳理了单库单目标防洪调度模型和梯级水库群多目标防洪调度模型,以水库群下游控制点的最大过流量最小和水库群最大泄流量最小为系统目标,考虑防洪控制点的重要程度和水库群的运行模式,建立了防洪调度管理优化多目标集成模型；并对上述模型运用快速非支配分类遗传算法(NSGA-Ⅱ)求解。实例计算表明,该模型和求解方法方便实用,可快速获得多组非劣解,构成非劣解集,不仅为复杂混联水库群联合防洪优化调度提供了一条有效的解决途径,而且也为防洪系统的调度管理决策提供充分的依据。(3)提出了能够有效解决混联水库群防洪调度管理中风险分析的方法。综合考虑水库群下游防洪安全和水库群本身的防洪风险,建立了混联水库群防洪调度管理风险分析模型；模型采用Monte Carlo模拟与“逼近于理想解的排序方法(TOPSIS)”相结合的方法进行求解,比较全面地描述了防洪系统的风险特征,为混联水库群防洪调度风险分析提供了一种有效的方法。(4)拓展了混联水库群防洪调度管理多目标决策理论与方法。论文首先论述了现有各种评价决策方法,结合混联水库群防洪调度的特点把信息熵方法引入权重计算,客观地反映方案集中具体数据对决策的贡献度,并利用可拓学在描述事物时能够将不同形式表述的因素组合并能够概括它们的全体特征的优点,建立了基于可拓学理论的防洪调度方案评价方法,为对水库群防洪调度管理可行方案集进行评估分析、寻求满意实施方案提供了有效的解决途径。(5)探讨了水库生态调度模型与方案优选。针对大型水库运行方式的特点,建立了兼顾社会、经济和环境效益的多目标优化运行模型,以期能提高水库的综合效益；同时,对排污权市场的制度建设进行了初步的设计,从交易的主体、交易的标的以及交易的规范管理等角度出发,分析了排污权交易市场的主要交易原则,结合水市场中的排污权交易实际,提出了一种基于协商模式的交易模型,并给出了各种情况下的优化解。(6)构建了混联水库群防洪调度管理决策支持系统。提出混联水库群防洪调度管理决策支持系统的总体组成,设计了内嵌模块——流域水文模型、防洪调度模型、数据处理模型等的主要功能。更多还原

【Abstract】 Reservoir, which is very important for regulating the discharge flow, cutting the flood peek, and alleviating the flood loss, is the effective engineering measures for development and management of river. In recent years, the water shortage and flood disaster usually take place in china. The reservoir ecological regulation which used for protecting the ecological environment is increasingly attracted the public’s attention. Nowadays flood control operation is difficult to meet the needs of all aspects. Therefore it is meaningful to study the management theory of flood control operation for mixed reservoirs for sustainable utilization of water resources, reducing the natural disaster and promoting the coordinated development of social, economic, and ecology. Based on the relationships of flood forecasting and the management of flood operation, optimization problems, risk analysis, scheme decision, ecological regulation, etc. of the mixed reservoirs are further discussed through intelligent algorithms, risk analysis theory, extension theory and other methods in this paper. The conclusions are as follows:(1) The flood forecasting techniques and parameter optimization method are improved. First this thesis analyses flood cause, hydrological model and parameter optimization method. In accordance with the above theory, flow calculus coefficient is optimized among the parameter estimation problem of Muskingum model by introducing the principle of minimizing the sum of absolute residuals. Considering the principle and the parameter characteristics of genetic algorithm, the Muskingum model parameter estimation adaptive accelerating genetic algorithm is presented in this paper.(2) Established the model and method of flood control operation for mixed reservoirs. This paper sums up the single-objective model of flood control operation for single reservoir and multi-objective model for cascade reservoir. Taking the important of flood control points and operation mode of cascade reservoir into account, Multi-objective integrated optimization model of flood operation management is found. The target of this is the maximum flow minimum of downstream control points and the maximum discharge flow minimum. NSGA-Ⅱis adopted to solve the above model. The effectiveness such as receiving multi-group non-inferior solution is verified by practical application. Therefore, this model provides a new idea and sufficient basis for flood control operation of mixed reservoirs.(3)The effective method to solve the risk analysis of the management of flood control operation for mixed reservoirs is proposed. Synthetically considering the downstream flood control safety and flood risk, management risk analysis model of flood control operation for mixed reservoirs is established. Monte Carlo simulation in combination with TOPSIS is used to solve the model and provides an effective method for risk analysis by describing the risk characteristics of the flood control system.(4) Developping the multi-objective decision theory and method for management of mixed reservoirs’flood control operation. Paper first discusses the existing variety of evaluation methods. Then, the information entropy is introduced to weight calculation. And this scheme objectively reflects the contribution for decision-making when the specific data is collected. Based on the extension theory, evaluation method for flood control operation is established. This method provides the novel assessment analysis opinion for management of flood control operation.(5) Discussing the ecological regulation and program optimization of mixed reservoirs. A multi-objective optimal operation model which is used to balance economic, social and environmental benefits is founded. It can also improve the efficiency and provide accurate and scientific information for the reservoir. At the same time, from the main body, the subject of the transaction, as well as trading standards and management point of view, a preliminary design about the river system is conducted. And the main trading principles are analyzed. Combined with the sewage in the water market trading practice, this chapter proposes a negotiation model which is based on the transaction model, and gives the optimal solution under various circumstances.(6) Constructing the management decision support system of flood control operation for mixed reservoirs. This paper designs the main function for embedded module which includes watershed hydrological model, flood control model, data processing model, etc.更多还原