基于电网运行数据集的电力系统运行评估及优化研究
发布时间：2018-11-02 10:15
【摘要】：随着智能电网不断发展,电力行业信息化程度不断提高,智能化元件设备不断应用到电力系统,电网自动化平台积累并处理了海量电力大数据。电力大数据具备大数据的基本特征,具有体量大、多样性、价值密度低及快速性特点,用传统数据方法难以及时、有效的处理。此外,电力大数据是电网运行过程中生产设备和监控设备产生的数据,具有明显时序特征。电力大数据的快速增长,给电力生产的计量、优化及调度工作带来的巨大的挑战。传统的数据处理方法限于数据处理能力,主要基于对整体信息进行采样,根据对采样数据确定电网典型运行方式,将电网典型运行方式分析结果进行外推,获得电网长期运行行为特性。而大数据的处理方法,可以直接以全体数据作为研究对象,利用数据推动的分析方法直接从数据中获得有效信息,进行电网运行状态表征及电网运行优化。本文提出了适用于电力大数据的数据预处理方法,利用大数据的冗余特点和电气量之间的物理关系,对缺陷数据进行修补,提高了电力大数据的利用效率。对电力大数据进行筛选,根据研究目标选取电网运行关键信息,构建电网运行数据集,为基于运行数据对电网进行分析提供了数据基础。提出了区分关键属性的部分优先聚类方法,对电网运行方式进行聚类。在此基础上求取电网典型运行方式,为电网优化研究提供有效的工具。通过提取电力大数据集每个时刻关键属性数据,生成对应时刻的电网运行方式特征参数数据集,并用聚类及聚类融合方法对该数据集进行聚类,可以得到电网典型运行方式及各典型运行方式出现概率,本文建立了基于电网运行数据集的电网有功网损评估模型,提出了电网有功网损评估方法。在此基础上建立了计及励磁系统调差系数的潮流计算模型,对发电机励磁系统调差系数进行优化,分析了不同调差系数方案下发电机对电网无功电压调节的影响,提出了基于电网数据集的发电机励磁系统调差系数优化整定方法,以提高电网电压水平,降低电网有功网损,可以充分考虑到电网发电、负荷等不确定性,使得优化结果更适于电网实际运行情况。
[Abstract]:With the development of smart grid and the improvement of power industry informatization, intelligent components and equipments have been applied to the power system, and the power system automation platform has accumulated and dealt with massive power big data. Power big data has the basic characteristics of big data, it has the characteristics of large volume, diversity, low value density and rapidity, so it is difficult to deal with it in time and effectively with traditional data methods. In addition, power big data is the data generated by production equipment and monitoring equipment in the operation of power grid, which has obvious timing characteristics. The rapid growth of power big data brings great challenges to the measurement, optimization and dispatch of power production. The traditional data processing method is limited to the data processing ability, mainly based on sampling the whole information, according to the sampling data to determine the typical operation mode of the grid, and extrapolating the analysis results of the typical operation mode of the power network. The long-term operation behavior characteristics of power grid are obtained. Big data's processing method can directly take the whole data as the research object, use the data-driven analysis method to obtain the effective information directly from the data, and carry on the power network operation state representation and the grid operation optimization. In this paper, a data preprocessing method suitable for power big data is proposed. The defect data can be repaired by taking advantage of the physical relationship between the redundant characteristics and electrical quantities of big data, and the utilization efficiency of power big data can be improved. In this paper, big data is screened, the key information of power grid operation is selected according to the research objective, and the data set of power grid operation is constructed, which provides a data basis for the analysis of power grid based on operation data. A partial priority clustering method for distinguishing key attributes is proposed to cluster the operation mode of power grid. On the basis of this, the typical operation mode of power grid is obtained, which provides an effective tool for the research of power network optimization. By extracting the key attribute data at each time of power big data set, the characteristic parameter data set of power grid operation mode is generated at the corresponding time, and the data set is clustered by clustering and clustering fusion method. The typical operation mode and the occurrence probability of each typical operation mode can be obtained. In this paper, an active power network loss evaluation model based on grid operation data set is established, and an evaluation method for active power network loss is proposed. On the basis of this, a power flow calculation model considering the adjustment coefficient of excitation system is established, and the adjustment coefficient of generator excitation system is optimized, and the influence of generator on the regulation of reactive power and voltage of power network under different schemes of adjustment coefficient is analyzed. Based on the data set of power grid, a method for optimizing the adjustment coefficient of generator excitation system is proposed to improve the voltage level of power network and reduce the loss of active power network. The uncertainty of power generation and load can be fully taken into account. The optimization results are more suitable for the actual operation of the power network.
【学位授予单位】：华北电力大学(北京)
【学位级别】：博士
【学位授予年份】：2017
【分类号】：TM732
[Abstract]:With the development of smart grid and the improvement of power industry informatization, intelligent components and equipments have been applied to the power system, and the power system automation platform has accumulated and dealt with massive power big data. Power big data has the basic characteristics of big data, it has the characteristics of large volume, diversity, low value density and rapidity, so it is difficult to deal with it in time and effectively with traditional data methods. In addition, power big data is the data generated by production equipment and monitoring equipment in the operation of power grid, which has obvious timing characteristics. The rapid growth of power big data brings great challenges to the measurement, optimization and dispatch of power production. The traditional data processing method is limited to the data processing ability, mainly based on sampling the whole information, according to the sampling data to determine the typical operation mode of the grid, and extrapolating the analysis results of the typical operation mode of the power network. The long-term operation behavior characteristics of power grid are obtained. Big data's processing method can directly take the whole data as the research object, use the data-driven analysis method to obtain the effective information directly from the data, and carry on the power network operation state representation and the grid operation optimization. In this paper, a data preprocessing method suitable for power big data is proposed. The defect data can be repaired by taking advantage of the physical relationship between the redundant characteristics and electrical quantities of big data, and the utilization efficiency of power big data can be improved. In this paper, big data is screened, the key information of power grid operation is selected according to the research objective, and the data set of power grid operation is constructed, which provides a data basis for the analysis of power grid based on operation data. A partial priority clustering method for distinguishing key attributes is proposed to cluster the operation mode of power grid. On the basis of this, the typical operation mode of power grid is obtained, which provides an effective tool for the research of power network optimization. By extracting the key attribute data at each time of power big data set, the characteristic parameter data set of power grid operation mode is generated at the corresponding time, and the data set is clustered by clustering and clustering fusion method. The typical operation mode and the occurrence probability of each typical operation mode can be obtained. In this paper, an active power network loss evaluation model based on grid operation data set is established, and an evaluation method for active power network loss is proposed. On the basis of this, a power flow calculation model considering the adjustment coefficient of excitation system is established, and the adjustment coefficient of generator excitation system is optimized, and the influence of generator on the regulation of reactive power and voltage of power network under different schemes of adjustment coefficient is analyzed. Based on the data set of power grid, a method for optimizing the adjustment coefficient of generator excitation system is proposed to improve the voltage level of power network and reduce the loss of active power network. The uncertainty of power generation and load can be fully taken into account. The optimization results are more suitable for the actual operation of the power network.
【学位授予单位】：华北电力大学(北京)
【学位级别】：博士
【学位授予年份】：2017
【分类号】：TM732
【相似文献】
相关期刊论文 前10条
1 ;电力行业电网运行与控制标准化技术委员会成立[J];电力建设;2000年08期
2 刘晓贤;;对银南电网开展“六复核”工作的探讨[J];宁夏电力;2004年03期
3 陈彦锋;;创新电网安全经济运行的对策[J];中国高新技术企业;2007年15期
4 林森;;2008年3月华东电网电网运行总体情况[J];华东电力;2008年04期
5 ;电网运行培训部[J];山东电力高等专科学校学报;2010年02期
6 吴敏;;冰冻威胁电网安全 广西电网调度中心积极防控[J];广西电业;2011年01期
7 姜瑞峰;;加强电网运行维护安全工作的措施[J];农村电气化;2013年06期
8 蒙定中;大电网与大机组的协调(一)[J];电力系统自动化;1986年05期
9 ;提高认识、狠抓落实，努力开创我局电网运行工作的新局面──张惠清总工程师在青海电网运行工作会议上的总结讲话[J];青海电力;1996年03期
10 诸兵;;浅谈电网运行技术的优化[J];科技与企业;2013年24期
相关会议论文 前10条
1 ;全国电网运行与控制标准化技术委员会成立[A];《电站信息》2009年第1期[C];2009年
2 闵勇;鲁宗相;乔颖;徐飞;;风电从“并网”到“融入”大电网运行的转变[A];2010年电力系统自动化学术研讨会论文集[C];2010年
3 韩自奋;;甘肃电网气象信息服务系统应用功能探讨[A];水电站梯级调度及自动控制技术研讨会论文集[C];2008年
4 吴琛;李玲芳;张丹;张杰;程e，
本文编号：2305735
本文链接：https://www.wllwen.com/kejilunwen/dianlidianqilunwen/2305735.html
