Tree-line Contradictory Risk Early Warning Technology of Rural Distribution Network Considering Severe Convective Weather

Abstract

Tree-line contradiction refers to the conflict between planting trees in the power line erection space and ensuring the safe and stable operation of power lines. In severe convective weather, the tree-line contradiction is further intensified. Once the tree-barrier grounding fault is triggered, the power supply reliability of the rural distribution network will be seriously affected. Aiming at this, a zonal early warning mechanism for tree-line contradiction of rural distribution networks considering severe convective weather is proposed in this paper. Firstly, due to the imbalance of tree-barrier grounding fault records, the Synthetic Minority Oversampling Technique (SMOTE) algorithm is used to replace part of the majority class samples with minority class samples, and the data preprocessing is realized based on keeping the scale of the data set unchanged. Secondly, combined with six meteorological monitoring indexes of severe convective weather, a mapping model of severe convective weather and tree barrier grounding risk is established based on Extreme Learning Machine (ELM). Finally, the data of a rural distribution network in China is used for example analysis. The results show that the proposed model can effectively reflect the mapping relationship between severe convective weather and tree-barrier grounding faults, and accurately realize the zonal warning of tree-barrier grounding risk, which has good robustness and scalability.