Authors: Shaoqing Shi; Zhuo Xu; Yong Xiao
Addresses: China Southern Power Grid, Guangzhou 510080, China ' China Southern Power Grid, Guangzhou 510080, China ' Electric Power Research Institute of China Southern Power Grid, Guangzhou 510080, China
Abstract: In order to ensure the safe and economic operation of power grid, a load identification method of household smart meters based on decision tree algorithm is proposed. This paper pre-processes the missing data, noise data and inconsistent data in the load data of household smart meter, and uses the decision tree algorithm to predict the load data after pre-processing. According to the prediction results, combined with mathematical tools, from the PQ characteristics, current characteristics, V-I characteristics The load characteristics of household smart meters are extracted from the characteristics, harmonic characteristics and instantaneous characteristics and the objective function of load identification is constructed based on the combination of characteristics, so as to realise the load identification of household smart meters based on decision tree algorithm. Comparative results show that this method can reduce the error rate of load, to improve the efficiency of identification, identifying the shortest time of only 1.5 s.
Keywords: decision tree; household smart meter; load data; feature combination.
International Journal of Global Energy Issues, 2022 Vol.44 No.5/6, pp.440 - 453
Received: 14 Jan 2021
Accepted: 12 May 2021
Published online: 08 Sep 2022 *