基于区间证据理论的多源融合广域后备保护
Multi Source Fusion Wide Area Backup Protection Based on Interval Evidence Theory
汪伟 1王吉文 1谭颖婕 2刘雯琳 3戴长春 1于洋 1罗国敏2
作者信息
- 1. 安徽电力调度控制中心,安徽合肥 230022
- 2. 北京交通大学电气工程学院,北京 100044
- 3. 国网湖北省电力有限公司鄂州供电公司,湖北鄂州 436000
- 折叠
摘要
针对传统保护方案在光伏接入系统时存在的可靠性不足问题,该文提出了一种基于区间证据理论的多源融合广域后备保护方案.首先,将光伏输出电流对保护的影响解耦分析,提出了保护方案的基本概率分配函数,将传统保护结果的0,1值转化为描述故障概率的区间值,定量考虑了光伏接入对配网保护的影响.其次,利用改进的区间证据理论对多种保护的判别结果进行融合,得到每条线路的综合故障概率区间,进而识别故障线路.该方法将光伏接入对保护判别的影响定量化,利用了多个保护综合判断,提高了广域后备保护的可靠性.同时,改进的区间证据理论消除了冲突证据的影响,在保护存在数据缺失及错误时仍能可靠识别故障线路.最后仿真验证了该文方法的可靠性.
Abstract
For insufficient reliability existing in traditional protection schemes in photovoltaic(PV)access systems,a multi-source fusion wide area backup protection scheme based on interval evidence theory is proposed.Firstly,the impact of the PV output current on protection is decoupled and analyzed,and a basic probability allocation function for the protection scheme is proposed.The values of 0 and 1 in traditional protection results are transformed into interval values that describe the fault probability so that the impact of the PV access on the distribution network protection can be quantitatively considered.Secondly,the improved interval evidence theory is used to fuse the discrimination results of multiple protections,obtaining the comprehensive fault probability interval for each line,and further identifying the faulty line.This method quantifies the impact of PV connection on protection discrimination and utilizes multiple comprehensive protection judgments to improve the reliability of the wide area backup protection.Meanwhile the improved interval evidence theory can eliminate the impact of conflicting evidence and can reliably identify faulty lines in the presence of data missing and errors.Finally,the reliability of the proposed method is verified through simulation.
关键词
光伏电源/10kV配电网/后备保护/区间证据理论/继电保护Key words
photovoltaic power supply/10 kV distribution network/backup protection/interval evidence theory/relay protection引用本文复制引用
基金项目
国家电网安徽省电力公司科技项目(52120021N00L)
国家自然科学基金(52077004)
出版年
2024
NETL
NSTL
万方数据