基于电器行为特征的非侵入式家用负荷识别

Non-intrusive Household Electrical Apparatus Identification Based on Electric Behavioral Characteristics

针对家用电器负荷识别中电气特性相似的电器难以识别的问题,提出基于电器行为特征的非侵入式负荷识别方法。该方法是在利用稳态电气特征的负荷印记进行家用电器负荷识别的基础之上,增加行为特征负荷印记;建立了负荷印记统计模型,对未知事件负荷印记的置信度进行模糊识别;提供识别结果以及识别结果的综合隶属度,提高了家用电器识别的精确度和准确度。经实验数据检验以及实际用户数据印证,该方法简单、准确、可靠,为非侵入式负荷提供新的思路和参考。

Aiming at the problem that it’s difcult to identify the d appliances load with the similar electrical characteristics in household load identification, a non-invasive load identification method based on electric behavioral characteristic is proposed. The method is based on the use of steady-state electrical characteristics, increasing the behavioral characteristics as the new load signatures for household load identifcation. The statistical model of the load signature is established, then the unknown event with the confdence of load signatures is identifed by fuzzy recognition. As a result, the identifed appliance with the comprehensive membership is provided, improving the accuracy of non-invasive household load identifcation. It is proved by the experimental data test and the actual user data that the method is simple, accurate and reliable, and providing new ideas and references for non-invasive load identifcation.