Decentralized Distributionally Robust Dispatch of Multi-Regional Power Systems Considering the Correlated Variable Wind Power

The power generation resources and load centers are not evenly distributed geographically in the power system and the proportion of wind power (WP) increases, which brings about the critical need for joint dispatch of multi-regional power systems (MRPS). In this article, the dispatch of MRPS considering the correlation of variable WP is formulated using the distributionally robust optimization in a decentralized way. We use the augmented Lagrangian duality to decompose the original problem into regional subproblems, which is solved by the alternating direction method of multipliers. To exploit the correlation of WP, the ambiguity set of WP is constructed using the principal component analysis. Comparative simulations show that the quality of solution obtained by the decentralized model is the same with that by the centralized model, and the involvement of the correlation of WP reduces the level of conservativeness.