Correlation Analysis Based Characteristic Identification Method for Harmonic Source

With the increasing number of nonlinear loads and the prevalence of users with harmonic sources connected to the power grid, harmonic pollution at the public common connection point (PCC) is worsening. Current power quality monitoring systems cannot perform specialized monitoring for each user, making it challenging to identify the respective harmonic characteristics of multiple users on the feeder. Therefore, a characteristic identification method for harmonic source based on correlation analysis is proposed. Firstly, large-capacity industrial users are screened based on user capacity. Secondly, temporal features such as typical production cycles and switching moments of working conditions are characterized. Then the time series data of multi-user power and power quality monitoring data at PCC are segmented based on the changing point detection, and the correlation between power and monitoring indexes for each time period is calculated to obtain the power quality characteristic of harmonic users. Finally, the feasibility of the proposed method is verified by the measured data.