Sub-second state estimation implementation and its evaluation with real data

This paper describes the performance of a parallel state estimation (PSE) software implemented using advanced computing techniques. The developed code is able to solve state estimation of large-size, practical power systems within one second. Benchmarking tests against a commercial Energy Management System tool shows that the computational speed is more than 10 times faster with same level of accuracy. Benefits brought by the sub-second state estimation are also discussed in the aspects of improving robustness of state estimation, maintaining power system reliability, and enhancing market efficiency.