An efficient data-driven approach for solving chance-constrained AC optimal power flow

Alternating current optimal power flow (ACOPF) economically dispatches controllable conventional generation subject to power system constraints such as power flow, generation, and transmission limits. With increasing renewable penetration, the inherent variability of these energy sources necessitates the modeling of uncertainty in the ACOPF, which is formulated as chance-constrained ACOPF (CC-ACOPF) in this paper. The CC-ACOPF is originally computationally challenging to solve, mainly due to the huge number of scenarios needed to model chance constraints and highly non-convex power flow. Therefore, this paper proposes a computationally tractable method for solving CC-ACOPF based on a scenario reduction technique and a datadriven convexification of non-convex AC power flow (ACPF). The main novelties of the proposed solution approach lie in leveraging 1) a fast algorithm for scenario technique (FAST) that does not rely on probability distribution and requires a smaller number of scenarios than the conventional scenariobased solution method (SSM) for modeling chance constraints and 2) a highly accurate data-driven convexification of ACPF. Eventually, the proposed approach reformulates the originally intractable CC-ACOPF model into a computationally tractable convex model. The simulation results on IEEE test cases show the efficiency and efficacy of the proposed solution approach in solving chance-constrained ACOPF.