Data-model hybrid-driven adaptive voltage control for active distribution networks

The increasing integration of renewable energy resources in active distribution networks (ADNs) aggravates voltage deviations. Fluctuations among distributed generators (DGs) and the absence of accurate network parameters hinder the performance of model-based voltage control method. How to utilize the advantages of measurement-based data-driven method combined with model-based method has become the key to effectively addressing voltage issues. This paper proposes a data-model hybrid-driven adaptive voltage control method for ADNs. A data-model hybrid-driven adaptive voltage control framework containing two hybrid modes is established with the consideration of measurement configuration. In the data-model correction mode, the performance of data-driven control is improved by prior physical knowledge in adequate measurement area. In the data-model coordination mode, the inter-area coordination realizes the complementarity of the regulating ability between the areas with adequate measurement and those without. Finally, analysis and verification are performed based on the modified IEEE 33-node distribution network. The results demonstrate that the proposed hybrid-driven voltage control method has superiority in adaptability to DG fluctuations and strategy interpretability, which obtains satisfied voltage control performance.