Abstract: Under the background of the increasingly prominent voltage sag problem in modern power grids，it

is of great engineering significance to accurately identify the users' profile of demand of voltage sag control for

guiding users' precise investment decisions for voltage sag control and helping power grid companies to identify the

objects of premium power supply value-added services. Based on the multi-attribute user data，taking into account

the users' demand of voltage sag control and the service willingness of the power companies，a voltage sag demand portrait label system including the users' voltage sag process characteristics and service characteristics was

established，which comprehensively describes the supply and demand relationship between power users and power

companies. Secondly，the fuzzy C-means clustering algorithm was used to make user group portraits in order to

realize the characterization and classification of user group characteristics. Then，based on the entropy weight

method，the comprehensive score of the users' two-dimensional indicators was calculated to describe the users'

voltage sag control demand and clarify the premium user group of power companies. Finally，based on the user data of 107 enterprises in the distribution network of a city in my country，the rationality and effectiveness of the method proposed were verified.

      Key words: voltage sag；users' profile；data-driven；demand identification