The application provides a kind of
electromagnetic field fast prediction method and
system of
maglev train suspension
electromagnet, for the coil strong source of suspension
electromagnet under typical working condition, nonlinear constitutive of iron core and local high gradient multiscale
coupling problem, based on
electromagnetic field sample
data extraction coil area and construct distance field, according to the structure characteristics of suspension
electromagnet and the distance field, the solving area of electromagnet is divided into coil conductor subdomain and multiple iron core subdomains of different distance levels, and
subnetwork is respectively constructed.In the training process, the residual error of Maxwell quasi-static equation and the consistency constraint of material constitutive are introduced in the subdomain, the magnetic
vector potential continuity and the normal continuity constraint of magnetic induction intensity are applied on the interface of subdomain;Further introduce the time-varying mixed scheduling mechanism of the right end source term of iron core subdomain, to improve the prediction stability and precision of strong
source area and high gradient area.In the reasoning stage, the
electromagnetic field distribution can be quickly output, and support is provided for fast calculation of electromagnetic force,
eddy current loss, etc.