基于提升算法的电力系统故障录波数据压缩新方案  

Novel Compression Scheme of Fault Recording Data in Power Systems Based on Lifting Algorithm

作　　者：乐全明[1] 郁惟镛[1] 柏传军[2] 费铭薇[1] 杜俊红 

机构地区：[1]上海交通大学电子信息与电气工程学院,上海市200240 [2]南京南瑞继保电气有限公司,江苏省南京市211100 [3]浙江省临安市供电局,浙江省临安市311300

出　　处：《电力系统自动化》2005年第5期74-78,共5页Automation of Electric Power Systems

摘　　要：利用提升算法实现了一个基于小波变换的电网故障录波数据压缩的新方案,从采样频率的角度首次给出确定最优分解层数的公式,说明了压缩比与频率的关系;结合实际录波数据,选定误差上限后,给出了通用阈值法中参数的优化取值;对比分析了软硬阈值处理方法。实例分析结果表明,利用提升算法实现的小波数据压缩算法比Mallat算法节省一半左右的计算量;利用文中提出的最优分解层数计算公式和优化参数,结合硬阈值法,可以获得较高的压缩比,同时将误差限制在合理的范围内。A novel scheme based on wavelet transform and its reversible transform of fault recording data compression in power systems by adopting lifting algorithm is presented in this paper. The computing formula to determine optimal decomposition layers is derived for the first time from the point of view of sampling rate. The relationship between compression ratio and sampling rate is further expounded. All the parameters can be optimized in the course of determining general threshold values when the error upper limit is given. The applicability of the soft-threshold and the hard-threshold method has also been compared. Practical example result demonstrates that the computing time of the lifting scheme is nearly the half of that of Mallat algorithm when applying them to the data compression by using the same wavelet transform. The higher compression ratio is obtained if adopting the proposed formula of the optimal decomposition layers, optimized parameters and the hard threshold method. In addition, the errors of compression can be controlled within a reasonable range.

关 键 词：小波 提升算法 数据压缩 分解层数 阈值 故障录波 优化参数 

分 类 号：TM711[电气工程—电力系统及自动化]