水电站水库群调度优化及其效益评价方法研究

发布时间：2018-06-07 05:47

本文选题：梯级水库群 + 支持向量机　；参考：《华北电力大学》2014年博士论文

【摘要】：能源是人类生存和发展的重要物质基础,攸关国计民生和国家安全。水电作为目前开发规模庞大、开发技术最为成熟的可再生能源,以其良好的调节性能、低廉的运行成本和快速的负荷响应能力,在世界电力能源格局中发挥着重要作用。我国水力资源丰富,为经济社会发展提供了能源保障。加快开发水能资源是我国增加清洁能源供应、优化能源结构、应对世界气候变化、实现可持续发展的重要措施。“十二五”时期是我国全面建设小康社会的关键时期,从我国的能源特点和自然资源结构来看,加快水电发展也是实现2020年非化石能源目标的必经之路,也是有效降低单位GDP二氧化碳排放量的重要措施。水库调度是水库运行管理的重要环节,调度水平直接影响着水库水电站综合效益的发挥。合理优化的水库调度方式能够在不增加硬件投入的情况下,获得可观的社会效益和经济效益,也是优化能源结构、促进节能减排的有效措施。本论文在对水库水电站群隐随机优化理论回顾归纳的基础上,分别从确定性优化调度模型建立与求解、调度规则的制定与优化、基于调度规则的水库水电站群系统仿真、效益评价及隐随机优化调度因素影响分析等方面对径流不确定条件下的水电站群优化调度进行研究。主要研究工作包括： (1)水库水电站群隐随机优化调度理论研究及归纳。介绍了水库确定性优化调度和随机调度的概念和特征,从径流过程角度分析二者之间的区别和关系。在分析显随机优化调度和隐随机优化调度原理的基础上,重点综述隐随机优化理论方法的国内外研究进展及其在水电站水库调度规则制定中的应用,并总结各种调度规则制定方法的适用条件和优缺点。 (2)基于网格搜索和交叉验证的改进支持向量机模型研究。基于支持向量机方法的原理分析其在回归预测领域的优势,针对支持向量机对参数敏感和小样本回归易受训练样本随机性影响的特点,建立基于网格搜索的参数寻优机制和基于交叉验证的样本随机性规避机制,对支持向量机性能进行改进。通过实例研究,验证改进机制对支持向量机在小样本训练拟合能力和预测能力方面的效果。 (3)基于C++和MATLAB的水库水电站群混合编程仿真平台的建立。针对隐随机优化调度在实际运行中的实现难度,考虑隐随机优化调度模型复杂、计算机实现环境多样化的特点,以支持向量机理论为例,将基于MATLAB的调度决策生成算法预测编译为动态库文件,使其在基于C++的水库水电站群系统仿真程序中被调用,实现实时滚动模拟。通过案例应用,对仿真平台的结构及系统稳定性和可扩展性进行评价。 (4)金沙江中下游12级梯级水电系统隐随机优化调度研究及其效益评价。以我国十三大水电梯级中规模最大的金沙江中下游梯级水电站系统为例,以系统发电量和保证出力为优化目标,建立并求解梯级中长期确定性优化调度模型,作为隐随机模型的训练样本。运用改进支持向量机方法对系统调度规则制定,并模拟系统1989-2000年运行过程。另基于多元逐步回归法制定调度规则并仿真,将同期确定性优化调度结果及两种仿真结果进行对比。对仿真结果的发电量、发电过程、保证出力等方面进行对比,分析仿真结果的效益和可靠性。 (5)隐随机优化调度模型因素影响研究。定量研究梯级规模、径流预报误差、模型参数、输出决策等因素对梯级水电站群隐随机优化调度仿真结果的影响。基于金沙江下游——长江中游大型梯级水电系统,以其宗单库、其宗——向家坝12级和其宗——葛洲坝14级三种电站组合为研究对象,控制各影响因素变化范围,并分别进行仿真运行和效益评价。评价结果所揭示的各因素所带来的影响方式对于支持向量机理论的改进以及隐随机优化调度的下一步发展有着重要的参考价值。
[Abstract]:Energy is an important material basis for the survival and development of human beings. It is vital to the national economy and the people's livelihood and national security. As a renewable energy, which has a large scale of development and the most mature development technology, it plays an important role in the world power energy pattern with its good regulation performance, low operating cost and rapid load response ability. China is rich in hydraulic resources and provides energy security for economic and social development. Speeding up the development of water energy resources is an important measure for China to increase the supply of clean energy, optimize the energy structure, cope with the world climate change and achieve sustainable development. "12th Five-Year" period is the key period for China to build a well-off society in an all-round way, from China's energy special. Point and natural resource structure, speeding up the development of hydropower is also the only way to achieve the goal of non fossil energy in 2020, and it is also an important measure to effectively reduce the emissions of GDP carbon dioxide.
Reservoir operation is an important link in the operation and management of the reservoir. The level of dispatching directly affects the comprehensive benefit of the reservoir. The rational and optimized reservoir scheduling method can obtain considerable social and economic benefits without increasing the input of hardware. It is also an effective measure to optimize the energy source structure and promote energy conservation and emission reduction. On the basis of the review of the theory of hidden stochastic optimization for reservoir hydroelectric stations, this paper is based on the establishment and solution of the deterministic optimal scheduling model, the formulation and optimization of the scheduling rules, the simulation of the reservoir hydroelectric station group system based on the scheduling rules, the benefit evaluation and the factor influence analysis of the implicit stochastic optimization scheduling, and so on. The optimal operation of hydropower stations is studied. The main research work includes:
(1) the study and induction of the implicit stochastic optimization scheduling theory of reservoir hydropower stations. The concept and characteristics of reservoir deterministic optimal scheduling and stochastic scheduling are introduced. The difference and relationship between the two are analyzed from the point of view of the runoff process. On the basis of the analysis of the explicit stochastic optimization scheduling and the implicit stochastic optimization scheduling, the implicit stochastic optimization theory is mainly summarized. The research progress of the method at home and abroad and its application in the formulation of hydropower station reservoir scheduling rules, and the application conditions and advantages and disadvantages of various scheduling rules formulation methods are summarized.
(2) an improved support vector machine model based on grid search and cross validation. Based on the principle of support vector machine (SVM), the advantages of the support vector machine in the domain of regression prediction are analyzed. In view of the characteristics of the parameter sensitivity of support vector machines and the randomness of the small sample regression which are easily subject to the randomness of the training samples, the parameter optimization mechanism and the base based on the grid search are established. The performance of SVM is improved by the random evasion mechanism of cross validation. The effect of the improved mechanism on the fitting ability and prediction ability of the support vector machine in small sample training is verified by an example.
(3) the establishment of a hybrid programming simulation platform for the reservoir hydropower station group based on C++ and MATLAB. In view of the difficulty of realizing the hidden random optimization scheduling in the actual operation, the characteristics of the complexity of the hidden stochastic optimization scheduling model and the diversification of the computer environment are considered, and the support vector machine theory is taken as an example, and the scheduling decision generation algorithm based on MATLAB is predicted. As a dynamic library file, it is called in the simulation program of the C++ based reservoir hydroelectric station group system. The real time rolling simulation is realized. The structure of the simulation platform, the stability and extensibility of the system are evaluated by the case application.
(4) the research and benefit evaluation of the cascade hydropower system in the middle and lower reaches of the middle and lower reaches of the Jinsha River, taking the cascade hydropower stations in the middle and lower reaches of the middle and lower Jinsha River, the largest in the thirteenth big hydropower cascade in China as an example, to establish and solve the middle and long term deterministic optimal scheduling model of the cascade. Training samples of implicit stochastic model. The system scheduling rules are formulated with improved support vector machine (improved SVM), and the 1989-2000 year operation process of the system is simulated. In addition, the scheduling rules are formulated and simulated based on the multiple stepwise regression method. The results of the deterministic optimal scheduling and the two simulation results are compared. Compare the process, guarantee output and other aspects, and analyze the effectiveness and reliability of the simulation results.
(5) study on the factor influence of the implicit stochastic optimization scheduling model. Quantitative study of the influence of cascade scale, runoff forecasting error, model parameters, output decision and other factors on the simulation results of cascade hydropower stations' implicit stochastic optimization scheduling. Based on the lower reaches of the Jinsha River, the large cascade hydropower system in the middle reaches of the Yangtze River, with its single library and its sect to Jiaba 12 level The combination of the three kinds of power stations in Gezhouba Dam 14 is the research object, which controls the range of the influence factors, and carries out the simulation operation and the benefit evaluation respectively. The influence mode of the factors revealed by the evaluation results has an important reference for the improvement of the support vector machine theory and the next step of the hidden stochastic optimization scheduling. Value.
【学位授予单位】：华北电力大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TV737;TV697.12

【参考文献】