Soft Open Point Planning Method for Active Distribution Network Based on Generative Adversarial Network

With the continuous increase of the penetration rate of distributed generation (DG) in the distribution network, higher requirements have been put forward for the dependable and cost-effective operation of the distribution network. In order to address the impact caused by the intermittent output of a large number of DGs on the distribution network, this article proposes a planning method for soft open point (SOP) using generative adversarial network (GAN). First, to accurately simulate the operating characteristics of DG, a GAN is employed for sce-nario generation. Then, operational constraints for distribution networks with SOP are proposed, with the objective function being the minimization of annual total cost. By relaxing and linearizing the original non-convex and non-linear constraints, a second-order cone programming model is derived, and the objective function is solved under the constraints. Finally, the proposed scenario generation algorithm and planning model are analyzed and veriffed on the IEEE 33- node test system.