基于分层匹配追踪算法的电能质量复合扰动参数辨识方法

发布时间：2018-03-18 06:43

本文选题：电能质量　切入点：原子分解　出处：《电力自动化设备》2017年03期 　论文类型：期刊论文

【摘要】：针对电能质量中的复合扰动信号分析问题,提出一种粒子群优化(PSO)和匹配追踪(MP)算法相结合的分层搜索的原子分解方法。首先应用MP算法提取基波分量,对于去除基波分量的残差信号,利用快速傅里叶变换找寻能量最大的频率成分,采用PSO算法粗搜索出最佳匹配粒子,然后以最佳匹配粒子为中心,在一定范围内重新离散化,生成小规模原子库,再应用MP算法有针对性地进行细搜索,最终得到最佳匹配原子,提取出电能质量复合扰动特征参数。仿真结果表明,该方法能克服MP算法匹配时间长、计算量大及PSO优化MP算法残差积累过大、容易陷入局部最优、匹配参数不准确等缺点,且具有一定的抗噪性和实时性。
[Abstract]:According to the power quality of the composite disturbance signal analysis problems, put forward a particle swarm optimization (PSO) and matching pursuit (MP) algorithm for hierarchical search combines the atomic decomposition method. First extracting the fundamental component using the MP algorithm, for the removal of residual signal is the fundamental component, the use of fast Fourier transform for maximum energy frequency components the PSO algorithm is used to search the optimal matching of coarse particles, and then the best matching particle as the center, in a certain range of re discretization, generating a small atomic library, then apply the MP algorithm to fine search, and ultimately get the best matching atom, extracting multiple power quality disturbances simulation parameters. The results show that the method can overcome the MP algorithm, a long time, a large amount of calculation and optimization of PSO MP algorithm for residual accumulation is too large, easy to fall into local optimum matching parameters are inaccurate, and has a certain The anti noise and real-time performance.

【作者单位】：燕山大学电气工程学院电力电子节能与传动控制河北省重点实验室;
【基金】：国家自然科学基金资助项目(51477147) 河北省教育厅资助科研项目(QN2015124) 燕山大学青年教师自主研究计划课题(1B14027)~~
【分类号】：TM711

【相似文献】