Speedup of EMT simulations by using an integration scheme enriched with a predictive Fourier coefficients estimator

Abstract

In this paper, we present a new version of the Sinusoidal Predictor Method, whose objective is to accelerate the EMT simulations of AC power systems. More precisely, it aims at using larger time steps to integrate the arising systems of differential algebraic equations. Contrary to main of the existing approaches, the SPM takes into account the sinusoidal behavior of oscillating components in the solver instead of the modeling part and so does not modify the original equations representing the simulated system. This methodology, which combines an estimation method with an adaptive step size integration scheme, allows us to use a wide range of step sizes varying from 1 μs in transients to 100 ms in steady-state. In this new version of the SPM, modifications have been provided to the estimation step: henceforth, the Fourier coefficients are computed by optimizing a measurement of the system stationarity. As a result, the number of iterations used for simulating a system is dramatically reduced compared to a standard integration scheme, especially in steady-state and during slow dynamics.