An Operational Risk Assessment Method Based on Decision Tree Regression for Distribution Network

With the expansion of the distribution network's size and the installment of numerous new energy sources, people pay more and more attention to the safe and reliable operation of distribution network. This article suggests a decision tree regression-based method for operational risk assessment for distribution networks. Firstly, the principal component analysis method is utilized to minimize the dimensionality of the distribution network operation data gathered by PMU and extract the key features. Secondly, a voltage stability index is proposed to evaluate the safety of real-time operation of distribution networks. Finally, the reduced-dimensional distribution network operation data is combined with voltage stability indicators and input into the decision tree regression algorithm for training. Based on the trained model, real-time voltage stability indicators of the power grid are predicted, achieving online operational risk assessment for distribution network. The proposed method is tested in IEEE 33-node distribution network model, which proves that it still has good evaluation performance under fewer PMU configurations and is suitable for online application.