摘要
基于固有时间尺度分解ITD算法,提出了改进ITD算法:针对波形失真问题,引入了有理样条函数插值和极值点延拓;针对虚假分量问题,增加了能量判据作为迭代终止条件。把改进ITD算法与边际谱熵相结合,用于判别高压输电线路单相接地故障类型。首先ITD分解两种故障相电弧电压获得边际谱,然后分析对比了两种故障边际谱的特点,从时频域定性揭示了两种故障相电弧电压及其边际谱的差异,再根据信息熵原理,得出ITD边际谱熵来快速准确识别两种故障类型。结果表明:输电线路瞬时性故障时,熵值大于等于0.2;永久性故障时,熵值小于0.2。针对输电线路单相接地故障的仿真结果,验证了该方法有效、可行。
An improved intrinsic time-scale decomposition（ITD）algorithm is presented：for waveform distortion,rational spline function interpolation and extrapolation at extreme points are introduced;for false weight,the energy criterion is added as the termination condition for iteration. The improved ITD algorithm and marginal spectrum entropy are combined to distinguish the single-phase ground fault types for high-voltage transmission line：first,ITD decomposes two faults with arc voltage to obtain the marginal spectrum,then the characteristics of marginal spectrum are compared, and the difference in the marginal spectrum of fault phase voltages from time-frequency domain is revealed;according to the information entropy principle,the ITD marginal spectrum entropy can distinguish between two fault types rapidly and accurately. The results show that when the transmission line is with transient fault,the entropy value is greater than 0.2;and with permanent fault,the entropy value is less than 0.2. The simulation results of transmission line with single-phase ground fault demonstrate that the proposed algorithm is effective and feasible.
出处
《电力系统及其自动化学报》
CSCD
北大核心
2016年第10期24-30,共7页
Proceedings of the CSU-EPSA
基金
国家高技术研究发展计划(863计划)资助项目(2012AA050215)
国家电网公司科技资助项目(5216A313500N)
关键词
单相自适应重合闸
固有时间尺度分解
边际谱熵
瞬时性故障
永久性故障
single-phase adaptive reclosure
intrinsic time-scale decomposition（ITD）
singular spectrum entropy
transient fault
permanent fault
