Controllable Photovoltaic Scenario Generation via Mixup-based Deep Generative Networks

As a type of Monte Carlo simulation, scenario generation is an effective method to solve the uncertainty problem in stochastic planning of integrated power systems. This paper proposes a novel model for photovoltaic (PV) scenario generation by employing interpretable condition in controllable generative adversarial networks (GANs). In order to improve the generalization performance of the network and increase the robustness to adversarial examples, a data augmentation strategy is introduced to the network. Simulation results demonstrate that, the proposed model can achieve the controllable generation of scenarios covering explicit statistical characteristics and produce brand new patterns not covered by the existing trajectories.