A comprehensive machine learning-based investigation for the index-value prediction of 2G HTS coated conductor tapes

• DOI: 10.1088/2632-2153/ad45b1
• Publication Date: 2024-04-30T22:43:39Z
• Authors: Shahin Alipour Bonab; Giacomo Russo; Antonio Morandi; Mohammad Yazdani-...
• Keywords: Value (mathematics)

Abstract Index-value, or so-called n- value prediction is of paramount importance for understanding the superconductors’ behaviour specially when modeling superconductors needed. This parameter dependent on several physical quantities including temperature, magnetic field’s density and orientation, affects high-temperature superconducting devices made out coated conductors in terms losses quench propagation. In this paper, a comprehensive analysis many machine learning (ML) methods estimating has been carried out. The results demonstrated that cascade forward neural network (CFNN) excels scope. Despite needing considerably higher training time compared to other attempted models, it performs at highest accuracy, with 0.48 root mean squared error (RMSE) 99.72% Pearson coefficient goodness fit ( R -squared). contrast, rigid regression method had worst predictions 4.92 RMSE 37.29% -squared. Also, random forest, boosting methods, simple feed can be considered as middle accuracy model faster than CFNN. findings study not only advance but also pave way applications further research ML plug-and-play codes studies devices.