Parallel power system transient stability analysis on hypercube multiprocessors

With the growth in size and complexity of interconnected power systems the simulation of power system equations in the time domain for stability assessment is becoming more time-consuming or too expensive. There is always considerable incentive to find better computation schemes. The authors develop an efficient and fast computation scheme for transient stability analysis on a hypercube multiprocessor. In order to realize communications among processing elements (PEs) efficiently, a pseudo-binary tree is embedded into the hypercube. The pseudo-binary tree is a binary tree structure where a node in the hypercube may correspond to more than one node in the corresponding binary tree. The implementation results on an iPSC/2 are discussed in detail to suggest an efficient way to parallelize the computation for power system transient stability analysis.<
>