复杂水库群供水优化调度方法及应用研究

发布时间：2018-04-02 01:26

本文选题：水库群供水调度　切入点：供水任务分配　出处：《大连理工大学》2016年博士论文

【摘要】：我国的大型江河流域已经或正在形成水库群联合开发利用总体布局,科学合理的水库群联合调度已经成为流域水资源优化配置与实时调控的关键,但随着水库群规模的增大水库群联合调度遇到新的难题,如复杂水库群中共同供水任务分配、水库群的优化调度求解、多目标优化调度决策等问题。本文以辽宁省东水西调中线工程为例,对上述三个问题从调度模型构建、模型优化求解、多目标决策制定三个层面开展研究,并对跨流域引水中人工引水的补偿作用机制进行探索性研究,主要研究成果有以下几个方面：(1)从社会经济发展状况、供用水情况、水资源特点等多个方面分析辽宁省东水济西的水资源配置格局,基于长系列径流资料、用水资料对中线水利工程的来水、用水资料进行分析,概化中线水利工程及用水户的拓扑结构,明确中线工程的调度难题及目标。(2)构建通用的水库群供水调度模型,确定调度规则的基本形式,将中线工程的原始调度目标转化为相应的目标函数,并给出约束条件及决策变量,最后结合现有的模型求解技术,分析其存在的问题并提出相应的解决策略。(3)针对水库群多目标联合调度中决策变量多、求解难度大的问题提出了基于敏感性分析的降维优化方法。首先利用全局敏感性分析方法来挖掘各决策变量及其之间相互作用对调度结果的影响,在此基础上建立仅包含敏感变量的简化模型,求解简化模型得到Pareto解集,并将其作为完整调度模型的初始解,以达到降维及提高搜索效率的目的,并以辽宁省东水西调工程为例对方法进行验证。结果表明：Sobol全局敏感性分析方法可用于识别水库群多目标联合优化调度问题中决策变量的敏感度,且基于该敏感性结果的降维优化方法提高了搜索效率。(4)针对复杂水库群共同供水任务分配的难题,提出了基于补给限制线的补偿调节法,改善已有补偿调节法不考虑补水水库自身供水任务而导致补偿不足或过度补偿的问题；在此基础上进一步提出综合考虑当前库容、水库入流以及水库供水压力的动态分水比例法。两种分配方法能够合理划分共同供水任务,有效发挥库容补偿作用,避免了水库供水的深度破坏,同时在保证整体最优的基础上均衡了三个子系统的供水目标,使三个子系统的供水质量更加均衡,保证子系统之间的公平性,具有明显的优越性。(5)多目标进化算法可以获得众多的非支配解,形成Pareto解集,但这给决策带来新的挑战,针对Pareto解集优选问题,提出一种基于改进拥挤度算子的权重组合设计方法,利用欧式距离来计算拥挤度,距离越远越不相似,进而保证权重组合的多样性；之后利用生成的权重组合从Pareto解集的968个解中选出13个解,为决策者提供一个决策表；最后基于公平性原则提供了推荐方案。(6)连通条件下的水库的人工引水可以对天然径流形成补偿,降低其不确定性,受水水库可以以较小的库容对径流进行调蓄,以大伙房应急入连工程中的碧流河水库为例,探索性研究人工引水对天然径流的补偿作用。结果表明：人工引水对天然径流的补偿作用呈现非线性关系,水库的缺水指数与人工引水之间呈现典型的倒S型曲线；在这条曲线上存在两个拐点,拐点出现的位置受用水户需水量与天然径流控制；人工引水的加入可以降低水库的库容而不减小水库的供水效益,在碧流河水库的实例中,当人工引水量达到天然径流的54%时,仅仅需要原始水库库容的60%即可对入流进行调蓄,满足用水户的保证率要求。
[Abstract]:Large river basins in China have been or are being formed reservoirs of exploitation and utilization of the overall layout, combined with scientific and reasonable scheduling of multi reservoir has become the key of basin water resources allocation optimization and real-time control, but with the increase of the size of reservoirs joint operation of reservoirs have encountered new problems, such as complex reservoirs in common water supply task assignment. Optimization of reservoir group scheduling problem of multi objective decision making. This paper takes Liaoning Province East West Water Transfer Project as an example, to the above three questions from the scheduling model, model optimization, multi-objective decision making in three aspects to carry out research, exploratory research and compensation mechanism of the artificial water cross in the basin water diversion, the main research results are as follows: (1) from the development of social economy, water supply and demand situation, many aspects of the characteristics of water resources analysis in East Liaoning Province The pattern of water resources allocation of water Jixi, a long series of runoff data based on the data of runoff water of the middle route of water conservancy projects, water use data analysis, generalized midline hydraulic engineering and the topological structure of the water diversion project scheduling problem, clear and objective. (2) the establishment of group scheduling model of general water supply reservoir, to determine the basic form the scheduling rules, the original scheduling objective of the middle route project into the corresponding objective function and constraint conditions are given, and the decision variables, finally combining model to solve the existing technology, analyzed the existing problems and puts forward corresponding solutions. (3) according to the decision of reservoirs multi-objective scheduling in multi variables, solving difficulty the proposed dimensionality reduction method based on sensitivity analysis. First, using the global sensitivity analysis method to explore the decision variables and the interaction between the impact of scheduling results On this basis, the simplified model contains only the sensitive variables, a simplified model is obtained Pareto solution set, and as a complete scheduling model of the initial solution, in order to achieve dimensionality reduction and improve the search efficiency, and to the East West Water Transfer Project in Liaoning Province as an example to verify the method. The results showed that: Sobol global the sensitivity analysis method can be used to identify the joint optimal scheduling decision variable reservoirs multi-objective problem in sensitivity, and improve the efficiency of search based on dimensionality reduction method of the sensitivity of the results. (4) to solve the problem of complex reservoir water supply joint task allocation, put forward the adjustment method of supply limit line compensation based on improving the existing compensation adjustment method does not consider the water reservoir own water supply task due to lack of compensation or over compensation; on this basis, considering the current capacity of reservoir inflow and water Dynamic water diversion ratio method. Two kinds of reservoir water pressure distribution method can divide common water supply, effective capacity compensation, avoid the reservoir water depth of damage, while ensuring the best overall balance on the basis of the three subsystems of water supply, the three subsystems of the water quality is more balanced, to ensure fairness among the subsystems, has obvious advantages. (5) the non dominated solutions of multi-objective evolutionary algorithm can obtain many, the formation of Pareto solution set, but the decision to bring new challenges for the Pareto solution set of optimization problems, put forward a design method for improving weight combination crowding operator based on calculation congestion by Euclidean distance farther from the similarity, so as to ensure the diversity of weight combination; after using the weight combination generated from the Pareto solution set of 968 solutions from 13 solutions for decision makers Provide a decision table; finally the recommended scheme is provided based on the principle of fairness. (6) the reservoir connectivity under the condition of artificial water can be formed to compensate the natural runoff, reduce the uncertainty of water reservoir on runoff for storage with a smaller capacity, emergency into even Biliuhe reservoir engineering for patients taking Dahuofang, compensation of the exploratory research on the artificial water runoff. The results show that the compensation effect of artificial water diversion on the runoff are nonlinear, inverted S curve between reservoir water index and artificial water diversion; there are two inflection points on this curve, the inflection point of location for users water demand and natural runoff control; artificial water could reduce the storage capacity of the reservoir without reducing the reservoir water supply efficiency, in the case of the Biliuhe reservoir, when the artificial water amount of natural runoff At 54%, only 60% of the reservoir capacity of the original reservoir can be adjusted for the inflow to meet the requirement of the user's guarantee rate.

【学位授予单位】：大连理工大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TV697.12

【相似文献】