基于云计算平台的统计综合法负荷建模的研究

发布时间：2018-03-23 17:32

本文选题：负荷建模　切入点：综合统计法　出处：《郑州大学》2017年硕士论文

【摘要】：电力系统负荷的数学模型是电力系统数字仿真的基石,而数学模型的精准程度会对仿真结果的准确性产生直接影响,从而影响到以此为依据的规划与决策。在过去的数十年中,发电机、励磁系统及原动系统的建模技术都得到了长足的发展。电力负荷却由于自身时变性、分散性、非线性等特性停滞不前,尤其是作为负荷建模三种方法之一的统计综合法已明显落后于时代。随着我国“全球能源互联网”战略的提出,电网复杂程度陡增,电网安全稳定运行与落后负荷模型之间的矛盾越发的不可调和。因此,在当今智能电网新形势下,建立符合我国电网的综合负荷模型是当前负荷建模研究领域的重中之重。本文依托智能电网大数据环境,以各种新型的数据采集装置采集到的海量数据为基础,高性能云计算平台和MapReduce算法为技术支撑,运用综合统计法进行负荷建模。全文从负荷的模型与结构入手,分别对常用的静态和动态负荷模型,相应负荷特性的集结方法,以及系统辨识理论进行了介绍。本文首次提出在智能电网环境下,利用现有系统实现对负荷组成数据进行获取,并设计了云平台下信息采集模块构建方法。采用时下流行的MapReduce算法,对采集到的数据进行挖掘、整合。MapReduce是一种为海量数据集和分布式并行计算而设计的编程方式。该方法从函数式编程中获得灵感,将一个数学函数作为计算单元,Map(映射)和Reduce(化简)是其核心内容。与传统方法相比,该方法能够大大加快数据处理速度,良好的容错性以及分布式部署的特点能够很好地解决电网结构复杂,电压等级多造成的数据获取困难等问题。此外,本文还开发了人机交互界面,为软件操作者带来了优良的软件应用体验。本文最后以获取的数据为实验对象,选用MapReduce算法和综合负荷动态模型,对静、动态模型分别建模。获得的静态负荷模型能够较好地反映变电站综合负荷的静态负荷特性,云计算平台能够快速地准确地完成大量负荷数据的汇集统计工作;获得的动态模型验证了在对电力负荷进行有效分类的前提下,采用改进的加权平均方法建立负荷模型的有效性。
[Abstract]:The mathematical model of the load of power system is the cornerstone of the digital simulation of power system, and the degree of precision of mathematical model will have a direct impact on the accuracy of the simulation results, which affects on planning and decision-making. The generator in the past few decades, and the modeling of excitation system and control system have made considerable progress. The power load is due to their own time, dispersion, stagnation of nonlinearity and especially as the statistical synthesis method of three kinds of load modeling method has significantly lagged behind the times. With the development of China's "global energy Internet strategy, the complexity of the power grid increased sharply, the contradiction between the model of safe and stable operation of power grid and backward load more and more irreconcilable. Therefore, in the new situation of the smart grid, establish the comprehensive load model of power system in China is the current research of load modeling. Priority among priorities domain. Based on the smart grid data environment, with all kinds of new data acquisition device to collect massive data based, high performance cloud computing platform and MapReduce algorithm for technical support, the use of integrated statistical method for the load modeling. The full text from the model and the structure of the static load, respectively, and the dynamic load model, load characteristics of the corresponding aggregation method, and system identification theory are introduced. This paper proposes for the first time in the smart grid environment, realize the acquisition of load data using the existing system, and designed a method to construct the information acquisition module under the cloud platform by using MapReduce algorithm. The popular, mining for acquisition the data integration,.MapReduce is a kind of massive data sets and distributed parallel computing and design programming. This method get inspiration from functional programming, A mathematical function as computing unit, Map (map) and Reduce (Simplified) is its core content. Compared with the traditional method, this method can greatly accelerate the speed of data processing, fault-tolerant and distributed deployment features can well solve the problems of complex structure of power grid, voltage level caused by the data difficult to obtain. In addition, this paper also developed a man-machine interface, brought experience excellent software application software. Finally, the operator to obtain the data as the experimental object, using MapReduce algorithm and load dynamic model of static and dynamic model, respectively. The modeling of static load model can reflect the static load the characteristics of substation load, cloud computing platform to load data collection statistical work quickly and accurately; dynamic model obtained is verified in the power load. On the premise of effective classification, an improved weighted average method is used to establish the validity of the load model.

【学位授予单位】：郑州大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM743

【相似文献】