This code shows a parallel (GPU-computing based) implementation of the NR load flow method, including:
- parallel formation of sparse admittance and Jacobian matrices
- computation of real and reactive power equations
- formation of the linear system of equations and
- solution of the resultant sparse linear system using a parallel implementation of an iterative Krylov subspace method (Preconditioned Stabilized Bi-conjugate Gradient method)