摘要
针对广义S变换(generalized S-transform,GST)的参数优化困难问题,提出一种GST参数优化方法并应用到电能质量扰动分类与扰动参数检测中.对基频点对应的参数r独立设置突出时域扰动特征,便于其他频点的参数优化专注于频域扰动,使优化后的广义S变换(optimized generalized S-transform,OGST)能同时表征较高精度的时域扰动和频域扰动信息.提出确定r的优化指标,使r的求取具有自适应性和理论依据.基于OGST的时频矩阵,提出4种扰动特征,并设计决策树分类器进行分类识别.同时实现对扰动起止时间、扰动幅值、谐波成分等扰动参数的检测.仿真数据及实测数据分析表明OGST抗干扰能力强、识别精度和检测精度高.
Aiming at the difficulties of optimizing parameter r for generalized S-transform（GST）, a parameter optimized method for generalized S-transform is proposed and applied to the classification and detection of the power quality disturbances. The value of r for the fundamental frequency is set independently so as to reveal the time disturbance, convenient to optimize the values of r for other frequencies on the concentration of frequency disturbance. So the time-frequency disturbances characteristics can be obtianed simultaneously in desireable resolution. The evaluating indicators for determining r are put forward so as to provide the theoretical basis to determine r adaptively. Based on the time-frequency matrix of this method, four disturbance characteristics are presented for classification by the decision tree classifier. The disturbance parameters such as starting and ending time, disturbance amplitude and harmonic components are further detected in this study. The experiment analysis and the application to real power engineering demonstrate that the method has good noise robustness and satisfactory accuracy.
出处
《中国科学：技术科学》
EI
CSCD
北大核心
2016年第6期593-601,共9页
Scientia Sinica(Technologica)
基金
国家自然科学基金(批准号:51475405)
河北省自然科学基金(批准号:F2015203413
F2015203392)
秦皇岛市科技计划(编号:201502A043)资助项目
关键词
优化广义S变换
参数优化
决策树
电能质量分析
optimized generalized S-transform
parameter optimization
decision tree
power quality analysis
