摘要
为了实现对低压电器用电的实时监测,该文提出一种采用电压-电流(V-I)轨迹作为负荷特征的非侵入式负荷监测(NILM)算法。文中首先基于物理意义提出了十个V-I轨迹特征,然后以视在总功率的变化作为低压电器用电状态变化事件的检测标准,随后通过平滑和插值算法来提取V-I轨迹,最后基于V-I轨迹特征采用支持向量机多分类算法进行负载识别。测试结果表明,该算法具有较好的负荷识别精度。
In order to realize real-time monitoring of low-voltage electrical appliances,a non-intrusive load monitoring(NILM)algorithm using voltage-current(V-I)trajectory as a load characteristic is proposed.Firstly,ten VI trajectory features are quantified based on physical meaning,and then the change of apparent total power is used as the detection standard of low-voltage electrical power state change event.Then,the VI trajectory is extracted by smoothing and interpolation algorithm,and finally support vector machine multi-classification algorithm is used for load identification based on the VI trajectory characteristics.The feature uses the support vector machine multi-classification algorithm for load identification.The test results show that the algorithm has higher recognition accuracy than other algorithms using load recognition.
作者
杨云瑞
黄宇魁
陈文浩
YANG Yunrui;HUANG Yukui;CHEN Wenhao(Guangdong Topway Network Co.,Ltd.,Foshan 528200;Foshan Bureau,Guangdong Power Grid Co.,Ltd.,Foshan 528200)
出处
《计算机与数字工程》
2020年第10期2353-2358,共6页
Computer & Digital Engineering
关键词
非侵入式负荷监测
V-I轨迹
负荷识别
non-intrusive load monitoring
V-I trajectory
load identification
