GradNet Models

These examples use our physics-constrained neural-network framework for dynamic modeling of saturable synchronous machines, including spatial harmonics [1]. The gradient network (GradNet) architecture is based on [2]. This folder contains measured datasets for two PM synchronous reluctance machines: a four-pole 5.6-kW machine (Baldor ECS101M0H7EF4) and a 10-pole 28-kW machine (Brusa HSM1.10.18.04). Additionally, a FEM dataset with spatial harmonics is provided for the 5.6-kW machine.

References

[1]

Li, Foissner, Martin, Piippo, Hinkkanen, “Gradient networks for universal magnetic modeling of synchronous machines,” 2026, https://arxiv.org/abs/2602.14947

[2]

Chaudhari, Pranav, Moura, “Gradient networks,” IEEE Trans. Signal Process., 2025, https://doi.org/10.1109/TSP.2024.3496692

5.6-kW PM-SyRM, train flux map, FEM data, no spatial harmonics

5.6-kW PM-SyRM, train flux map, FEM data, no spatial harmonics

5.6-kW PM-SyRM, train current map, FEM data with spatial harmonics

5.6-kW PM-SyRM, train current map, FEM data with spatial harmonics

5.6-kW PM-SyRM, GradNet from FEM data, FVC

5.6-kW PM-SyRM, GradNet from FEM data, FVC

5.6-kW PM-SyRM, GradNet from measured data, FVC

5.6-kW PM-SyRM, GradNet from measured data, FVC

28-kW PM-SyRM, train flux map, measured data

28-kW PM-SyRM, train flux map, measured data

5.6-kW PM-SyRM, train flux map, measured data

5.6-kW PM-SyRM, train flux map, measured data

5.6-kW PM-SyRM, train current map, measured data

5.6-kW PM-SyRM, train current map, measured data

28-kW PM-SyRM, train current map, measured data

28-kW PM-SyRM, train current map, measured data

5.6-kW PM-SyRM, train flux map, FEM data with spatial harmonics

5.6-kW PM-SyRM, train flux map, FEM data with spatial harmonics