摘要
基于奇异谱提出的奇异熵概念用于对次同步振荡模态的定阶,采用与矩阵束算法结合的奇异熵矩阵束算法对电力系统的次同步振荡参数进行辨识。首先选取加入噪声的理想信号对Prony算法与奇异熵矩阵束算法的有效性、辨识精度、最小频率间隔辨识值及辨识所需最小数据量等辨识能力进行比较分析;然后分别利用两种算法对IEEE第1标准测试系统及某实际串补输电工程模型进行进一步分析验证。分析结果表明,奇异熵矩阵束算法具有有效性,可以方便、准确地确定模态阶数,提高频率分辨率,降低所需数据量,而且具有很强的抗噪能力和较高的辨识精度。
This paper based on the concept of singular spectrum is applied to determine the order of subsynchronous os-cillation modes,and singular entropy matrix pencil algorithm combined with matrix pencil algorithm to identify parame-ters of synchronous oscillation on power system identification. First the added noise of the ideal signal is selected to car-ry out comparative analysis of Prony algorithm and singular entropy matrix pencil algorithm in the aspect of validity,ac-curacy,minimum frequency interval identification value and the minimum amount of data needed for identification,etc. Then the IEEE first benchmark system and the actual series compensation project are further analyzed and verifiedby using two algorithms. The analysis results show that the singular entropy matrix algorithm is effective,can be conve-niently and accurately determine the modal order number,improve frequency resolution,reduce the amount of data re-quired,and have strong antinoise ability and higher recognition accuracy.
出处
《电力系统及其自动化学报》
CSCD
北大核心
2016年第6期31-36,共6页
Proceedings of the CSU-EPSA
基金
国家自然科学基金资助项目(51147001 51477010)
北京市重点学科--检测技术与自动化装置学科建设项目资助(5111523302)
关键词
次同步振荡
奇异熵
矩阵束算法
模态阶数
模态参数辨识
subsynchronous oscillation(SSO)
singular entropy
matrix pencil algorithm
modal order
modal parameter identification