Two-stage and multi-stage stochastic unit commitment under wind generation uncertainty

Although wind power is considered green energy and free, its intermittent nature causes the instability issues for the current power grid system. The two-stage stochastic optimization approach has been explored recently and justified as an effective approach to achieve cost efficiency while ensuring system reliability. This approach can fit well for most independent system operators (ISOs) today, performing reliability unit commitment (RUC) runs to ensure the sufficient thermal generation capacity to accommodate the wind output fluctuation. In this paper, we propose a multistage stochastic optimization unit commitment model and compare its performance with the two-stage stochastic optimization model, in terms of total expected cost and computational complexity. For the multi-stage stochastic optimization model, the unit commitment and economic dispatch decisions are made sequentially as more information is realized, which can take advantage of more realized information and provide more flexibility for unit commitment decisions to accommodate the uncertainty. The stochasticity of wind power output is represented by a scenario tree and the final computational results verify the value of the multi-stage stochastic optimization approach.