Meta-learning based voltage control strategy for emergency faults of active distribution networks

With the increase of energy demand and the continuous development of renewable energy technology, active distribution networks have become increasingly important. However, the introduction of a large amount of renewable energy has made the structure of ADN increasingly complex and fragile, and emergency fault caused by emergencies may often occur. Voltage control in the emergency fault event is particularly important. In this context, this paper presents a meta-learning based voltage control strategy for renewable energy integrated active distribution network. A general regression neural network is first applied to extract features from the operation data. Then, the local cross-channel interaction network is adopted to capture targeted information that is most related to emergency fault from the features and induce knowledge transfer to update the voltage control strategy. This allows the proposed strategy to make optimal decisions quickly when only limited data are available under an emergency fault that has never occurred. Comparison results based on a 69-bus distribution network validate the effectiveness and robustness of the proposed strategy.