基于电子电路拓扑结构模型驱动的非侵入式负荷识别

为区分相似负载之间的微妙差异,提出了一种家庭能源管理系统的非侵入式负荷识别方法,即利用模型驱动分类法的概念,重点研究了电源电路拓扑结构和多种常用电器的分类方法,前者是从电力电子角度提出分级分类方法,而后者则保证所提分类法在实际应用中的代表性。根据多层级分类方法,讨论了模型驱动分层特征提取的优点,结果表明该方法可获得90%以上的识别精度,从而为区分相似负载之间的微妙差异提供一种更简单、更可行的解决方案。

Non-intrusive Load Identification Based on Electronic Circuit Topology Model Analysis

In order to distinguish the subtle differences between similarity loads, a non-intrusive load identification method for home energy management systems is proposed. Using the concept of model-driven classification, this paper focuses on the topology structure of power supply circuits and the classification methods of many common electrical appliances. The former proposes a hierarchical classification method from the perspective of power electronics, while the latter guarantees that the proposed classification can be used in practical applications. According to the multi-level classification method, the advantages of model-driven hierarchical feature extraction are discussed. The results show that the method can obtain more than 90% recognition accuracy, which provides a simpler and more feasible way to distinguish the subtle differences between similar loads.