An efficient and robust eigenvalue method for small-signal stability assessment in parallel computers

Small signal stability analysis of interconnected power systems is performed through the partial computation of the eigenvalues/eigenvectors of very large matrices. Parallel processing has already been efficiently used in this analysis. In this paper, a parallel version of the bi-iteration method and of a new hybrid method specially developed for parallel processing are introduced and their performance compared with the results obtained with the parallel lop-sided simultaneous iterations algorithm. These algorithms were implemented in an Intel iPSC/860 parallel computer with hypercubic topology. The results presented in the paper were obtained for a model of a large practical power system.<
>