Fast parallel algorithm on sparse matrix for power system analysis

AbstractPower system analysis is the basis of operation and planning. Since power systems are very large and complex, however, it takes a long computation time for the analysis. It is desired to develop a fast method especially for transient stability analysis. In this paper, a parallel algorithm for sparse matrix factorization, forward substitution, and backward substitution is discussed. The algorithm is made operational on the actual transputer network. The computation time using the algorithm is measured and compared with that of the ordinary (single processor) algorithm. Finally, the parallel transient stability program is developed and evaluated.