Research on thermal control elimination method of screening current of high-temperature superconducting magnets based on K-means clustering algorithm

This paper proposes an improved heating elimination method to mitigate screening currents in high-temperature superconducting (HTS) magnets. To address the non-uniform target temperatures across different regions of the magnet during heating, a spatially non-uniform heating strategy is proposed to enhance its application in large and medium-sized magnets. First, an electromagnetic-thermal coupling model for HTS magnetic energy storage system is established, and the desired heating temperatures at key locations are back-calculated using the critical current formula of superconductors. The K-means algorithm is then employed to cluster the calculated temperature data, with the desired heating temperature as a one-dimensional feature. The clustering results are used to define heating zones and identify the temperature feedback points for each zone. Subsequently, fuzzy PID control is implemented based on the information from these temperature feedback points. Simulation results demonstrate that the combination of the K-means algorithm with fuzzy PID heating control can reduce the screening currents by 86.21%.