Periodic steady state solution of power systems by selective transition matrix identification and graphic processing units

Modern power systems analysis becomes more challenging due to their increasing complexity, greater interConnectivity of elements and sub-systems and the use of alternative generation systems, among others. In order to efficiently conduct these analyses, it is necessary the use of modern computational and numerical techniques, such as parallel processing and fast periodic steady state solution techniques. The Numerical Differentiation (ND) method has been successfully applied to meet the last issue. This paper introduces a modification to the ND method using GPUs, which allows a significant reduction on the computation time required to reach the steady state solution in nonlinear power systems.