基于贝叶斯准则的非侵入式负荷监测方法

Non-Intrusive Load Monitoring Based on Bayes Criterion

利用非侵入式负荷监测技术,既可以使居民了解家中电器详细的能耗信息,也有助于电力公司制定科学的需求响应政策。针对现有负荷分解方法采样频率要求高、无法有效处理多工作模式负载投入使用的问题,提出基于贝叶斯准则的负荷分解方法。首先介绍了一套用电信息采集系统;然后提出了负载电流概率密度函数的概念,并在此基础上建立了基于贝叶斯准则的负荷分解模型;最后给出了一种全新的非侵入式能耗估算方法。使用实测数据进行算法验证,结果表明所提方法可以精确估算出各家用电器的能耗。

Non-intrusive load monitoring（NILM） techniques can not only inform residents of energy consumption of household appliances but help electric enterprises formulate scientific demand response（DR） policy. In line with strict requirement on sampling frequency of the existing load decomposition method an failure of load operation with multiple running modes, this paper a load decomposition method based on Bayes criterion.Firstly, a power consumption information collection system is put forward. Then, probability density function of load current is introduced, and a Bayes criterion based load decomposition model is presented. Finally, a novel non-intrusive energy consumption calculation method is proposed. The proposed method is validated by the actually measured data that the proposed method can accurately figure out energy consumption of household appliances.