AC Loss Database Built with Numerical Multi-scale Model and Status Prediction of a 150 kJ SMES

Abstract

In the process of dynamic power compensation of superconducting magnetic energy storage system (SMES), AC loss is produced inevitably, which will influence the thermal stability of the SMES. In this paper, we firstly use a fast-numerical model, multi-scale model, to analyze the AC loss of a 150 kJ high temperature superconducting (HTS) SMES, and arrange the AC loss under different working conditions into a database. Then, based on the AC loss database, we build a neural network model to provide a real-time prediction of AC loss and thus to adjust the cooling power accordingly. This AC loss data base and neural network system will help keep the SMES magnet thermally stable and prevent the SMES from overload working condition.