当前位置:主页 > 科技论文 > 电力论文 >

基于广域信息的同调机群识别及电网失步解列新方案

发布时间:2018-11-16 14:23
【摘要】:大区域互联电网规模的扩大对系统的动态安全稳定问题的影响日趋严重,系统出现故障而引起的失步及连锁跳闸会对整个电网造成不可估量的损失。为提高和健全电力系统安全稳定性,完善“三道防线”的建设迫在眉睫。解列是保证系统安全运行的重要措施,作为保障电网安全稳定运行的最后一道防线,可以有效地防止事故的进一步扩大。而同调分群作为解列控制的重要前提,作用体现在简化动态行为,令系统迅速恢复同步。近年来通信技术的不断革新以及相量测量单元(Phasor Measurement Unit,PMU)的实际应用,为同调分群和解列等本已较成熟的课题的研究注入了新的生机和活力,在大区域电网互联的复杂背景下涌现出了大量的实践突破和大胆创新。基于广域测量系统(Wide Area Information System,WAMS),本文重点对以下两方面进行深入的探究:(1)提出了一种快速、有效的同调分群新方法,实现在线进行发电机分群的同时,克服了计算量大、实现时间长等缺点;(2)提出了一种新的解列方案,不受系统运行方式变化和结构改变的影响。将离散Fréchet距离应用于发动机组的在线分群,源数据为WAMS中实测的发电机功角。将采样功角值三角函数拟合面积差预分群后,再利用离散Fréchet距离达到在线分群的最终结果。与现有分群方法相比,该方法不受模型参数影响,适用于不同故障场景,尤其适用于复杂大系统,能够实时对系统进行动态分群。分析了系统失步时两侧电势幅值不等及系统运行方式变化对振荡中心和失步中心的影响。构造了一种能够反映振荡中心随功角摆开漂移的振荡中心位置函数,在此基础上提出了一种新的解列方案。该方案利用振荡中心位置函数的变化率辅助ucos?启动解列装置,克服了实际系统中电势幅值不相等情况下基于ucos?判据存在的缺陷。再综合振荡中心的漂移特点、失步中心的位置和解列后孤岛不平衡度来搜索解列断面。该方案能够捕捉振荡中心的变化轨迹和失步中心的位置。在选择和速动性得到保证的同时提高了解列装置的可靠性。
[Abstract]:The problem of dynamic security and stability is becoming more and more serious due to the expansion of the scale of large area interconnected power network. The loss of step and tripping caused by the failure of the system will cause incalculable losses to the whole power network. In order to improve and improve the safety and stability of power system, it is urgent to perfect the construction of "three lines of defense". De-listing is an important measure to ensure the safe operation of the system. As the last line of defense to ensure the safe and stable operation of the power grid, it can effectively prevent the further expansion of the accident. As an important premise of decoupling control, homology cluster can simplify the dynamic behavior and make the system restore synchronization quickly. In recent years, the continuous innovation of communication technology and the practical application of phasor measurement unit (Phasor Measurement Unit,PMU) have injected new vitality and vitality into the research of more mature subjects such as homology, grouping and uniting. A large number of practical breakthroughs and bold innovations have emerged under the complex background of large area power grid interconnection. Based on the wide-area measurement system (Wide Area Information System,WAMS), this paper focuses on the following two aspects: (1) A fast and effective homology clustering method is proposed to realize the on-line generator clustering at the same time. It overcomes the shortcomings of large computation and long time of realization. (2) A new de-listing scheme is proposed, which is not affected by the change of system operation mode and structure. The discrete Fr 茅 chet distance is applied to the on-line clustering of the engine group. The source data is the generator power angle measured in WAMS. After the sampling work angle trigonometric function is used to fit the area difference pregroup, the final result of on-line clustering is obtained by using the discrete Fr 茅 chet distance. Compared with existing clustering methods, this method is not affected by model parameters, and is suitable for different fault scenarios, especially for complex large scale systems, and it can dynamically cluster the system in real time. The influence of the amplitude difference of the two sides potential and the operation mode of the system on the oscillation center and the out-of-step center are analyzed when the system is out of step. A position function of the oscillation center which can reflect the shift of the oscillation center with the angle of work is constructed. On the basis of the function, a new scheme is proposed to solve the problem. This scheme uses the rate of variation of the oscillatory center position function to assist ucos? Start the disassembly device, overcome the potential amplitude in the actual system is not equal based on ucos? The defect of the criterion. Combining the drift characteristics of the oscillation center, the location of the out-of-step center and the unbalance degree of the island behind the train are used to search the discontinuous section. The scheme can capture the changing track of the oscillation center and the position of the out-of-step center. The reliability of the column device is improved while selection and rapidity are guaranteed.
【学位授予单位】:天津大学
【学位级别】:硕士
【学位授予年份】:2014
【分类号】:TM712

【参考文献】

相关期刊论文 前10条

1 刘志雄;张艳霞;冯康恒;马桦岩;李婷;;基于广域信息跟踪电力系统振荡中心的方法及应用[J];电网技术;2014年06期

2 程云峰;张欣然;陆超;;广域测量技术在电力系统中的应用研究进展[J];电力系统保护与控制;2014年04期

3 宋洪磊;吴俊勇;冀鲁豫;;基于慢同调理论和希尔伯特-黄变换的发电机在线同调识别[J];电力自动化设备;2013年08期

4 吴麟琳;黄少锋;;基于频率特征的失步解列判据原理[J];电力自动化设备;2013年01期

5 黄少锋;吴麟琳;;基于电气量频率差异的电力系统振荡识别[J];电力系统自动化;2012年23期

6 吴为;汤涌;孙华东;徐式蕴;;基于广域量测信息的电力系统暂态稳定研究综述[J];电网技术;2012年09期

7 张园园;龚庆武;陈道君;刘栋;;应用改进粒子群优化的模糊均值聚类算法的暂态稳定机组分群方法[J];电网技术;2011年09期

8 范瑾;;电力系统失步解列研究综述[J];西安邮电学院学报;2011年03期

9 张保会;汪成根;郝治国;;电力系统失步解列存在的问题及需要开展的研究[J];电力自动化设备;2010年10期

10 刘绚;文俊;刘天琪;;基于自组织神经网络的模糊聚类同调机群识别[J];电网技术;2010年07期

相关硕士学位论文 前5条

1 陈涛;基于PMU的同调机群识别及电网暂态稳定性预测新方法[D];华北电力大学(北京);2009年

2 高晓磊;基于离散Fréchet距的在线手写签名认证算法研究[D];武汉理工大学;2008年

3 董绍光;基于PMU的系统失步解列研究[D];山东大学;2007年

4 何孝军;电力系统动态等值中同调机群聚类方法研究[D];河海大学;2007年

5 李海燕;基于WAMS的电力系统暂态稳定预测方法研究[D];华北电力大学(北京);2006年



本文编号:2335758

资料下载
论文发表

本文链接:https://www.wllwen.com/kejilunwen/dianlilw/2335758.html


Copyright(c)文论论文网All Rights Reserved | 网站地图 |

版权申明:资料由用户b50b6***提供,本站仅收录摘要或目录,作者需要删除请E-mail邮箱bigeng88@qq.com