Conditional WGAN-GP-based Renewable Energy Scenario Generation Method

As the proportion of renewable energy in the power system continues to increase, the uncertainty of its output poses huge challenges to power system planning and dispatch. A renewable energy scenario generation method based on the Wasserstein GAN with Gradient Penalty (WGAN-GP) is proposed. WGAN-GP effectively solves the problem of unstable training in traditional Generative Adversarial Networks (GAN) by introducing conditional information, and can generate more diverse and representative scenarios. Verification was carried out using the GEFCom 2014 dataset. The results show that this method shows superior performance in terms of accuracy, autocorrelation and comparison with traditional methods such as Markov chains. This research provides a new solution for the generation of renewable energy scenarios and has broad application prospects.