Relaxed approach for the parallel solution of security-constrained dispatch with post-contingency rescheduling

Some approaches for the solution of the security-constrained dispatch with post-contingency corrective rescheduling problem in a multiprocessing environment are presented. These approaches can be interpreted as relaxed versions of the classical Benders decomposition. The idea is to improve the computational efficiency when the problem is solved by using parallel processing. Results from one of these approaches implemented into an asynchronous algorithm are reported here. Some strategies for improving performance are discussed and test results with two real-life Brazilian systems are reported.