A heuristic algorithm for maximum LOLP constrained unit commitment of wind power integrated system

This paper proposes a novel heuristic algorithm for maximum loss-of-load probability (LOLP) constrained unit commitment (UC) of significant wind power integrated system. Wind generation fluctuation and generator stochastic outage are considered in the model, and the maximum LOLP constrained UC is to prepare appropriate spinning reserve (SR) to keep the probability of loss of load under a predetermined upper limit. In the model, wind generation is treated as a multi-state unit according to its probability density function (PDF) which can be gotten by statistical or forecast approach. The algorithm solves the model by an iterative process between the traditional SR constrained UC and operating reliability estimation. Lagrangian relaxation (LR) is used here to solve the traditional UC problem while capacity outage probability table (COPT) is used to calculate the LOLP indicator. The specific SR update strategy makes the proposed heuristic algorithm more efficient. Test results on the RTS-96 system illustrate the efficiency of the proposed algorithm.