Machine learning approaches for identifying and predicting voltage conditions in power system networks using network topology behavior input formulation

The growing integration of fast-fluctuating energy resources poses potential challenges to power system operation. However, the widespread deployment of sensors across distribution networks enhances system visibility and enables the development of advanced voltage control strategies. These control methods depend on accurate voltage condition predictions to anticipate violation scenarios. This paper developed a predictive model for determining the condition of voltage in a power system based on the network topology behavior as input formulation. Different machine learning algorithms random forest, support vector machine and gradient boosting were used to evaluate the input formulation approach for voltage condition prediction in power system networks. In addition, various performance metrics were used to evaluate the model; the Matthews Correlation Coefficient (MCC) was also used to assess the predictive model's performance to avoid the effect of the imbalance in the dataset on the predictive model. The performance of the model indicates that the input formulation based on the network topology behavior in the power system is suitable for predicting the condition of voltage in the network. The MCC indicates that the models give an accuracy of 0.98 for the random forest model and 0.87 for both support vector machine and gradient boosting, respectively. The high accuracy of the model in identifying and predicting the voltage condition enables the operators to initiate necessary control actions to regulate the voltage within the acceptable limits in the network. In addition, the implementation of the predictive model to determine the voltage condition in the power system will ensure reliability and security in the power system network. Finally, the predictive model should be implemented with the control algorithim to increase the accuracy of voltage regulation in the power system.