High-current air-core inductor design optimization

This project aims to establish an optimized design framework for air-core inductors intended for the power converters used in particle accelerator systems. The methodology integrates analytical modeling, finite element analysis (FEA), prediction of internal electromagnetic forces, and artificial neural network (ANN) surrogate models, all embedded within a multi-objective optimization algorithm (targeting inductance, losses, mechanical stress, and compactness).

The methodology has been validated on a 100 kA pulsed inductor for accelerator applications, demonstrating accurate prediction of inductance and internal forces. The approach enables easy integration in the existing industrial design processes and environments, yielding a significant reduction in computational time and improved convergence in the design space.