摘要
给出了通用的相邻系数法表达式,并将其用于局部放电的去噪研究。与多小波传统阈值去噪算法相比,相邻系数法能较好地解决变换系数之间的相关性,获得更好的去噪效果。通过大量的仿真数据得出了均方误差与相邻系数邻域及指数之间的关系曲线,并在此基础之上得到了合适的相邻系数邻域及指数数值。现场数据的处理结果验证了多小波相邻系数法及选取的参数的有效性。
A universal expression of neighboring coefficient method is proposed and applied in denoising of partial discharge. Compared with traditional threshold denoising based on multiwavelet, the neighboring coefficient method can deal with the correlativity among the transformation coefficients and better denoising effect can be obtained. By means of a large amount of simulation data, the relation curve among mean square error, neighborhood of neighboring coefficient and index of neighboring coefficient is derived, on this basis appropriate value of neighborhood of neighboring coefficient and index of neighboring coefficient are obtained. The effectiveness of multiwavelet neighboring coefficient method and the chosen parameters is verified by the processing results of on-site data.
出处
《电网技术》
EI
CSCD
北大核心
2005年第15期61-64,70,共5页
Power System Technology
关键词
多小波
相邻系数
局部放电
白噪声
高电压绝缘技术
电力系统
Multiwavelet
Neighboring coefficient
Partial discharge
White noises
High voltage insulation engineering
Power system