Multi-objective unit commitment under hybrid uncertainties: A data-driven approach

In recent years, the growing penetration of renewable energy has increased the level of uncertainty in power systems, which brings challenges to modern unit commitment. This paper develops a data-driven unit commitment model with multi-objectives under wind power and load uncertainties. In particular, the distribution of the above uncertainties are estimated by a non-parameter kernel density method whose bandwidth is optimized to get more reliable and cost-effective UC solutions. To solve the complicated model, a reinforcement learning-based multi-objective particle swarm optimization algorithm is proposed. Finally, several experiments were carried out to demonstrate the effectiveness of this research.