基于稳态电流波形匹配的非侵入式家用负荷快速识别方法研究

Non-intrusive Household Load Rapid Identification Method Based on Steady-state Current Waveform Matching

针对现有非侵入式负荷识别方法存在识别特征量表征性差(选取不典型),识别算法收敛速度慢,识别结果精度低等问题,提出一种利用波形相似度匹配来辨识家用负荷的方法。将互近似熵原理与二值距离矩阵相结合,提取家用负荷稳态运行时的电流波形作为特征量,通过直接计算待识别的测试电流波形与模板库中的样本波形之间的互近似熵值,来进行波形相似度匹配,从而实现了对家用负荷类型的快速准确识别,并通过仿真实验模拟了家用负荷识别的过程。实验结果表明,该方法可以有效提高非侵入式家用负荷辨识的精度和效率。

Aiming at the problems of the existing non-intrusive load identification methods with poor characteristics of identification features(untypical selection),slow convergence speed of identification algorithms,and low accuracy of identification results,a method of using improved cross approximate entropy algorithm to identify household loads is proposed.Combine the principle of cross approximate entropy with the binary distance matrix,extract the current waveform of the household load in steady state ope-ration as the characteristic quantity,and directly calculating the cross approximation entropy value between the test current waveform to be identified and the sample waveform in the template library to perform the waveform Similarity matching,so as to realize the rapid and accurate identification of household load types,and the process of household load recognition is simulated through simulation experiments.Experimental results show that this method can effectively improve the accuracy and efficiency of non-intrusive household load identification.