Closed-Loop Preventive Dispatch for Resilient Power Systems in Abnormal Weather Events

Power systems are increasingly threatened by abnormal weather conditions, which are exacerbated by global warming. In particular, the proliferation of renewable energy sources might cause power systems to have insufficient energy flow in real-time to cope with the risk of severe power imbalances in extreme conditions. In this paper, a resilience-oriented closed-loop preventive dispatch model is proposed for power systems under abnormal weather conditions. Compared to the conventional open-loop dispatch model which relies on accurate forecasting, the proposed close-loop dispatch model can feed back resilience-oriented physical knowledge to ensure system resilience even in the case of insufficient forecasting. More specifically, the proposed model consists of three layers including a forecasting layer based on Bayesian neural networks, a stochastic power dispatch layer, and an assessment layer. A Bayesian backpropagation mechanism based on Karush-Kuhn-Tucker conditions is then developed to integrate the optimization problem in the architecture of deep neural networks. Finally, the results of numerical experiments have verified the effectiveness and superiority of the proposed model.