非侵入式负荷识别中的小波去噪分析

Analysis on the de-noising by using wavelet method for non-intrusive load identification

非侵入式负荷识别可以实现深入分析用户内部负荷成分,获取用户的负荷信息,完善用电信息采集系统与智能用电系统,支撑起双向互动服务和智能用电服务。从抑制噪声信号,精确提取电器的电气特性信号的方面,提出了采用小波变换对电流数据进行去噪的方法,通过理论分析和实验验证,均证明"coif1"具有良好效果,既能有效地抑制噪声信号又能完整地保存有效信号成份,在事件检测中,准确地提取出电器的电流波形信号,为非侵入式负荷识别研究提供了参考。

Non-invasive load identification can be carried out in-depth analysis of the user＇ s internal load components, and access to the user＇ s load information, improve the electricity information collection system and intelligent power system, and support from the two-way interactive services and intelligent electricity services. From the aspect of denosing and accurately extracting the electrical characteristic signal of the electrical appliance, a method of denoising the collected current data by wavelet transform is proposed. Through the theoretical analysis and experimental verification, it is proved that “eoifl” has good effect, which can effectively suppress the noise signal and completely preserve the effective signal components, in event detection, accurately extract the current waveform of the appliance and load recognition, and provided a reference for the study of the non-invasive load identification.