The changes in the pattern of electricity energy supply and demand, as well as the trend of energy development, have led to the reform and progress of the domestic and international electricity industry market, accelerating the construction of smart grids in various countries. In recent years, the rapid growth of China's economy has driven the continuous progress of science and technology. Information Technology (IT), Big Data (BD) technology and Artificial Intelligence (AI) are widely used in all walks of life. The traditional power grid is gradually developing towards user demand side response, making system regulation more flexible and allowing users to fully participate in the operation of the power system and energy market. In order to build a power grid user side intelligent management system based on AI algorithm driven by power BD, this article studies a user Demand Side Management (DSM) method based on load transfer technology and uses a demand side user response potential assessment model based on fuzzy optimization set to evaluate its management effectiveness. Through numerical simulation, the task allocation process of load aggregators is simulated, and the rationality, effectiveness, and feasibility of this assessment method are demonstrated from aspects such as task completion and aggregator benefits. The results show that the system can effectively control and manage the operation of various devices, keep the total consumption below the threshold, and manage the load based on priority.