Voltage sag sensitive load type identification based on power quality monitoring data

Abstract

This paper focuses on identifying voltage sag-sensitive loads within unknown load type or when the load is already in operation. To achieve this, a new method of sensitive load identification based on power quality monitoring data is proposed. Firstly, the active power RMS monitoring data is used as the base data. The Hodrick-Prescott filtering and sliding mean segmentation are used to divide the period of the voltage sag event. Next, based on the division result, the differences of steady-state power quality monitoring data before and after each event are calculated as the dataset. The dynamic K-means is used to divide various load action areas. Finally, the voltage tolerance curves of each action area are fitted and compared with the preset curves, then according to the constituted rules in this paper, the type of sensitive load contained by user is recognized. The feasibility and accuracy of the proposed method are verified by analyzing the simulation examples and actual power quality monitoring data.© 2017 Elsevier Inc. All rights reserved.

Keywords

• Sensitive load
• Power quality monitoring data
• Voltage sag
• Load identification
• Machine learning