EMT Simulation with Spectral Graph Wavelets

Electromagnetic transient (EMT) simulations provide a precise method to assess stability and simulate disturbances on electrical networks. Due to a rise in Inverter Based Resources (IBR), simulations require a smaller simulation time step to accurately model the high frequency controls in power electronics. This severely reduces the ability of system operators to study system dynamics with large electrical models, where simulations can occur upwards of days. This paper introduces a novel wavelet framework to reduce time-scale issues in co-simulation studies. A general method of multi time-step spatial partitioning is developed to address convergence issues in co-simulations. The proposed method leverages the Spectral Graph Wavelet Transformation (SGWT) to produce a low-resolution response to a high-resolution localized disturbance. The spectral transformations are used to approximate wave propagation in the external system, reducing the total simulation time and improving convergence.