基于多尺度线调频基稀疏信号分解的电力系统非平稳振荡信号特征提取
发布时间：2018-08-22 20:31
【摘要】：随着电力电子装置在源-网-荷侧的大规模应用,电力系统功率振荡表现出较强的非平稳特性和模式耦合性。本文提出一种基于多尺度线调频基稀疏信号分解的电力系统非平稳振荡信号特征提取方法。多尺度线调频基稀疏信号分解方法能够在动态的时间支撑区内对信号进行投影分解,逐次获得能量最大的信号分量,具有良好的时频聚集性,特别适用于非平稳振荡信号的分解。电力系统的功率振荡信号本质上是一种多模态时变振动系统响应信号。首先通过多尺度线调频基稀疏信号分解方法得到多个单模态振动响应信号,然后根据单模态振动响应特性采用最小二乘法进行振荡特征的提取。仿真算例和实例证明了该方法在电力系统非平稳振荡信号特征提取中的有效性和适应性。
[Abstract]:With the large-scale application of power electronic devices in source-network-load side, the power oscillation of power system shows strong non-stationary characteristics and mode coupling. In this paper, a method for feature extraction of non-stationary oscillation signals in power system based on multi-scale linear frequency modulation basis sparse signal decomposition is proposed. The sparse signal decomposition method based on multi-scale linear frequency modulation (MFM) can be used to decompose the signal in the dynamic time support region and obtain the signal component with the largest energy one by one, which has good time-frequency aggregation. It is especially suitable for the decomposition of nonstationary oscillatory signals. The power oscillation signal of power system is essentially a multi-mode time-varying vibration system response signal. Firstly, multiple single mode vibration response signals are obtained by the sparse signal decomposition method of multi-scale linear frequency modulation basis, and then the oscillation characteristics are extracted by using the least square method according to the single mode vibration response characteristics. Simulation examples and examples show the effectiveness and adaptability of this method in the feature extraction of non-stationary oscillation signals in power system.
【作者单位】： 东北电力大学电气工程学院;国网吉林省电力有限公司吉林供电公司;
【基金】：国家重点研发计划(2016YFB0900100) 国家自然科学基金(51377017)资助项目
【分类号】：TM712
本文编号：2198216
[Abstract]:With the large-scale application of power electronic devices in source-network-load side, the power oscillation of power system shows strong non-stationary characteristics and mode coupling. In this paper, a method for feature extraction of non-stationary oscillation signals in power system based on multi-scale linear frequency modulation basis sparse signal decomposition is proposed. The sparse signal decomposition method based on multi-scale linear frequency modulation (MFM) can be used to decompose the signal in the dynamic time support region and obtain the signal component with the largest energy one by one, which has good time-frequency aggregation. It is especially suitable for the decomposition of nonstationary oscillatory signals. The power oscillation signal of power system is essentially a multi-mode time-varying vibration system response signal. Firstly, multiple single mode vibration response signals are obtained by the sparse signal decomposition method of multi-scale linear frequency modulation basis, and then the oscillation characteristics are extracted by using the least square method according to the single mode vibration response characteristics. Simulation examples and examples show the effectiveness and adaptability of this method in the feature extraction of non-stationary oscillation signals in power system.
【作者单位】： 东北电力大学电气工程学院;国网吉林省电力有限公司吉林供电公司;
【基金】：国家重点研发计划(2016YFB0900100) 国家自然科学基金(51377017)资助项目
【分类号】：TM712
【相似文献】
相关期刊论文 前3条
1 任凌志;于德介;彭富强;皮维;;基于多尺度线调频基稀疏信号分解的齿轮故障信号阶比分析[J];机械工程学报;2011年13期
2 彭富强;于德介;刘坚;;一种基于多尺度线调频基的稀疏信号分解方法[J];振动工程学报;2010年03期
3 ;[J];;年期
相关博士学位论文 前2条
1 彭富强;多尺度线调频基稀疏信号分解及其在齿轮箱故障诊断中的应用[D];湖南大学;2010年
2 刘小峰;振动信号非平稳特征的深层提取技术及远程诊断服务系统的研究[D];重庆大学;2007年
相关硕士学位论文 前1条
1 陈关宝;基于多尺度线调频基稀疏信号分解的模态参数识别方法研究[D];湖南大学;2012年
，本文编号：2198216
本文链接：https://www.wllwen.com/kejilunwen/dianlidianqilunwen/2198216.html
