基于非侵入式用电数据分解的自适应特征库构建与负荷辨识

非侵入负荷监测是实现需求侧测量与能效优化的有效途径。文中提出了一种高频采集模式下的非侵入式负荷在线监测方法,使整个监测过程自动化、实时化。首先,根据负荷电流的可加性原理建立了负荷分离模型,得到独立负荷波形;并结合负荷的操作特性,无需预实验获取先验数据。然后,通过贝叶斯分类模型实现负荷种类判断,从而在运行过程中为每个独立用户构建动态的负荷特征库。最后,基于库中数据,通过构建寻优模型实现负荷辨识,从而持续、实时获取负荷用电状态,并通过实际采集的用电数据验证了方法的有效性。该研究可自适应地为独立用户构建负荷特征库,改善了提前建库不具有普适性的问题,同时,基于特征库的快速寻优保证了辨识的有效性与准确性。

Construction of Adaptive Feature Library and Load Identification Based on Decomposition of Non-intrusive Power Consumption Data

Non-intrusive load monitoring (NILM) is an effective way to realize measurement on demand side and optimization of energy efficiency. This paper explores an online NILM method in high-frequency acquisition mode, which ensures the entire process automative and real-time. Firstly, a load-decomposition model is established based on the additivity of load current to obtain the independent load waveform. Moreover, combined with the operation characteristics of load, priori data are obtained without pre-experiment. Then, the load types are judged by Bayesian classification model to construct dynamic feature library of load for every independent user during operation process. Finally, load identification is realized by constructing optimization model to continuously get the state of power consumption of load in real time. The power consumption data measured in actual scenario is used to verify the effectiveness of the method. The method can adaptively construct the dynamic feature library of load for independent users, which improves the weak universality caused by establishing the database in advance. The fast optimization based on feature library ensures the effectiveness and accuracy of identification.