Free-Shape Optimization of VHF Air-Core Inductors using a Constraint-Aware Genetic Algorithm

This paper focuses on the optimization of air-core inductors which are widely used in Very High Frequency (VHF) integrated converters. Instead of considering classical geometries (e.g., spiral and solenoid), a free-shape optimization algorithm is implemented, i.e. any geometry respecting the design rules can be considered. The optimization is performed with a fast Partial Element Equivalent Circuit (PEEC) solver and a custom genetic algorithm that enforces the non-linear design constraints. Finally, the inductor of a 1.6 W Integrated Voltage Regulator (IVR) operated at 40.68 MHz is optimized under various conditions (minimization of the losses, placement of the terminals, footprint constraint, and magnetic near-field reduction). It is found that the shape optimizer is particularly useful for problems with complex and/or unusual constraints.