【作者】 王洋；

【导师】 韩璞；

【作者基本信息】 华北电力大学 ， 控制理论与控制工程， 2011， 硕士

【摘要】 火电厂机组负荷优化分配,对于提高电厂的安全性、可靠性以及经济性具有重要的意义。本文首先分析了火电厂负荷分配的发展情况,在考虑火电厂机组实际运行特性的基础上,找到了影响火电厂机组间负荷优化分配的主要因素,实现了煤耗特性曲线的拟合,建立了各种负荷分配的数学模型。在讨论以标准煤耗量为主要目标的负荷分配中,本文在分析了各种单目标优化方法的优缺点的基础上,着重介绍了以等微增率法和动态规划法为代表的传统单目标优化算法。详细介绍了这两种方法的产生、发展以及在电厂负荷分配中的应用,设计了以这两种方法为基础的负荷分配方案。同时,随着电力市场的发展,火电厂负荷分配不仅需要满足负荷分配的经济性要求,还要满足负荷调整时间要求,以及在电力生产中的污染物排放要求。而以传统的方法仅仅考虑发电成本已经不能满足更高的负荷分配要求,因此,采用多目标方法对火电厂负荷分配进行研究具有重要的意义。相对于传统多目标算法,基于Pareto最优解的多目标智能算法具有明显的优势。针对这一课题,本文设计了两种多目标进化算法解决多目标负荷分配。分别是改进非支配排序遗传算法(NSGA-Ⅱ)和基于Pareto解集的多目标粒子群优化算法(PAMOPSO)。通过算例表明,本文的方法能很好地解决多目标优化问题,同时,将本文设计的多目标进化方法与加权方法的仿真结果进行了比较,证实了NSGA-Ⅱ和PAMOPSO的正确性和优越性,实例证明本文的多目标算法具有很好的研究意义和实际应用价值。更多还原

【Abstract】 Power Plant Units load dispatch, for improved plant safety, reliability and economy is of great significance. This article analyzes the development of thermal power plant load dispatch firstly, then, Consider the thermal power plant’s actual operating characteristics. Found the main factors which impact of thermal power unit load dispatch, realized coal consumption curve fitting and established a variety of mathematical models of load dispatch.In the discussion of the coal is the main objective of the load dispatch, the paper analyzes the advantages and disadvantages of various single-objective allocation method. At the same time, highlighting the equal incremental method and dynamic programming, represented by the traditional single-objective optimization algorithm. Introduced the emergence, development and Applications in load dispatch of the two methods in detail. Designed load dispatch program based on these two methods.At the same time, as the develop of electricity market, Load dispatch of thermal power plant is now not only to meet the economy requirement, but also adjust the time requirements and pollutant emissions requirements in the electricity production. However, the traditional method which only considers the higher cost of power generation can not meet the requirements of the load dispatch. Therefore, using multi-objective approach to study the load dispatch has important significance.Compared with the traditional multi-objective methods, the multi-objective evolutionary strategies based on the concept of Pareto optimal have obvious advantages. In response to this issue, we designed two multi-objective evolutionary algorithms to solve multi-objective load dispatch, NSGA-Ⅱand MOPSO which based on Pareto solution set, respectively. The numerical examples show that the method can solve multi-objective optimization problem very well .Meanwhile, we compared the paper design of the multi-objective evolutionary methods with the weighting, the simulation results show that the NSGA-Ⅱand PAMOPSO are superiority and accuracy, it also indicated that the algorithm has good practical application and significance.更多还原