Sensitivity analysis and machine learning modelling for the output characteristics of rotary HTS flux pumps

Abstract High temperature superconducting (HTS) rotatory flux pump, or so called HTS dynamo, can output none-zero time-averaged DC voltage and charge the rest of circuit if a closed loop has been formed. This type pump is often employed together with coils, where coils potentially work in persistent current mode, act like electromagnets considerable magnetic field, having wide range applications industry. The characteristics rotary pumps have extensively explored through experiments finite element method (FEM) simulations, yet on constructing statistical models as an alternative approach to capture key not studied published. A 2D FEM program used model evaluate effects different factors upon parameter sweeping analysis variance. Typical design considerations, including operation frequency, air gap, tape width remanent density investigated, particular bilateral effect explained by looking at averaged integration electric field over tape. Based data obtained from various regression conducted collection machine learning methods demonstrated that be promptly satisfactory accuracy via Gaussian process regression, aiming provide novel for future research powerful tool industrial using devices.