随机降维映射稀疏表示的电能质量扰动多分类研究  

Study on classification method of power quality disturbances based on random dimensionality reduction projection and sparse representation

作　　者：沈跃[1] 刘国海[1] 刘慧[1] 

机构地区：[1]江苏大学电气信息工程学院,镇江212013

出　　处：《仪器仪表学报》2011年第6期1371-1376,共6页Chinese Journal of Scientific Instrument

基　　金：国家自然科学基金(60874014);江苏省高校自然科学基金(10KJB470003);江苏大学高级人才启动基金(10JDG136)资助项目

摘　　要：提出一种随机降维映射特征提取与稀疏表示分类相结合的电能质量扰动信号识别方法。首先将扰动信号测试样本表示为训练样本集的过完备字典稀疏线性组合,然后使用随机测量矩阵获取测试样本降维特征量和稀疏表示感知矩阵,应用最小L1范数解决方案求取扰动信号测试样本的稀疏解,由冗余误差最小值确定目标归属类,实现对电能质量扰动的稀疏表示多分类识别。研究表明随机矩阵降维映射特征提取不依赖于电能扰动样本特性,构造简单,运算快速,具有普适性;稀疏表示分类法与支持向量机相比无需组合多个二分类器来实现多分类器。仿真和实验结果表明该方法能有效提取各种电能扰动特征,抗噪声鲁棒性好,在信噪比20 dB以上的噪声环境中电能质量扰动分类准确率达95%以上。A new approach combining random dimensionality reduction projection（RDRP） with sparse representation classification（SRC） is proposed to classify power quality disturbances（PQD）.At first,test sample is represented as a sparse linear combination of training samples in an over complete dictionary.Then,the dimensionality reduction feature vector of the test sample and sensing matrix of sparse representation are achieved using random matrix lower-dimensional projection.L1-minimization solution is applied to obtain the sparse representation of the test samples of PQD,the object class is determined through minimizing the residual error between test sample and its sparse representation.The sparse representation classification of PQD is realized.The random matrix dimensionality reduction projection feature extraction method is extremely efficient,simple and fast,and is independent of the sample characteristics of PQD.Compared with support vector machine（SVM）,the SRC algorithm doesn＇t need to combine two-class classifiers for multiclass classification.Simulation and experiment results show that the proposed RDRP-based SRC method has a high classification accuracy rate above 95% under 20dB noisy circumstance.

关 键 词：电能质量 扰动分类 压缩感知 随机矩阵 降维映射 稀疏表示分类 最小L1范数 

分 类 号：TM712[电气工程—电力系统及自动化]