Monitoring the charging behavior of electric vehicle (EV) clusters is helpful in developing more effective energy management strategies for grid operators. Non-intrusive monitoring for EVs has a comprehensive application prospect due to its low implementation cost. Aiming at the problem that traditional non-intrusive monitoring methods cannot identify unknown devices accurately due to the lack of classes, a non-intrusive monitoring method based on zero-shot learning (ZSL) is proposed in this paper, which can monitor the unknown types of EVs connected to charging piles. Firstly, the charging characteristics of known EVs and unknown EVs are extracted by dictionary learning. Then EVs is classified by ZSL based on sparse coding. Furthermore, EVs is decomposed based on the proposed multimode factorial hidden markov model (FHMM). Finally, the EVs dataset of Pecan Street is used to verify effectiveness and accuracy of the proposed method.

中文翻译：

---

基于零镜头学习的电动汽车非侵入式监控

监控电动汽车 (EV) 集群的充电行为有助于为电网运营商制定更有效的能源管理策略。电动汽车的非侵入式监测由于其实施成本低而具有广泛的应用前景。【摘要】：针对传统非侵入式监控方法由于缺少类而无法准确识别未知设备的问题,本文提出了一种基于零样本学习(ZSL)的非侵入式监控方法,可以对未知类型进行监控。连接到充电桩的电动汽车。首先，通过字典学习提取已知电动汽车和未知电动汽车的充电特性。然后通过基于稀疏编码的 ZSL 对 EV 进行分类。此外，基于提出的多模因子隐马尔可夫模型（FHMM）分解电动汽车。最后，