Multifrontal Solver for Online Power System Time-Domain Simulation

This paper proposes the application of unsymmetric multifrontal method to solve the differential algebraic equations (DAE) encountered in the power system dynamic simulation. The proposed method achieves great computational efficiency as compared to the conventional Gaussian elimination methods and other linear sparse solvers due to the inherent parallel hierarchy present in the multifrontal methods. Multifrontal methods transform or reorganize the task of factorizing a large sparse matrix into a sequence of partial factorization of smaller dense frontal matrices which utilize the efficient Basic linear algebra subprograms 3 (BLAS 3) for dense matrix kernels. The proposed method is compared with the full Gaussian elimination methods and other direct sparse solvers on test systems and the results are reported.