摘要
针对目前电能质量扰动定位与识别困难的问题,提出了基于多特征量的电能质量复合扰动识别方法。该方法将EEMD、改进TK能量算子、Hilbert变换、扩展Prony算法相结合得到信号主要特征量,确定各特征量阈值,设计了决策树分类器进行快速的扰动识别,避免了因训练样本不足引起的较大误差,在较大程度上缩短了识别时间。选取包括10种复合扰动在内的17种扰动信号进行仿真验证,仿真试验结果表明,该方法识别率高,抗噪能力强,可同时适用于单一和复合电能质量扰动信号的识别。
Aiming at the difficult positioning and recognition of the electric energy quality compound disturbance,the recognition method of composite power quality disturbance based on multiple characteristics is proposed.By combining the ensemble empirical mode decomposition （EEMD）method, the improved TK energy operator, the Hilbert transform and the extended Prony algorithm,main characteristics of the signal are obtained ,and the threshold value of each characteristic is determined.A decision tree classifier is designed for fast disturbance identification,which avoids the large error caused by the lack of training samples, and shorten the recognition time in a great extent.17 kinds of power quality disturbance signals,which including 10 kinds of composite disturbances are used in simulation.Simulation results showed that the method has high recognition rate ,strong anti-noise ability, and can be applied to the signal recognition of both single and composite power quality disturbances.
出处
《山东电力技术》
2017年第4期16-21,共6页
Shandong Electric Power
