Prediction of Flashover Voltage on Contaminated Insulators Using Machine Learning

Abstract

Electrical insulators are a crucial component of electrical power transmission systems, and the issue of insulator pollution is one of the most significant challenges facing electrical networks. Pollution leads to electrical breakdowns, causing power outages and substantial financial losses. Therefore, accurately predicting the Flashover voltage of polluted insulators is essential for improving insulator design and ensuring high operational efficiency and reliability of power systems. insulator design and ensuring high operational efficiency and reliability of power systems. The use of artificial intelligence techniques significantly contributes to enhancing the accuracy of predicting the Flashover voltage of polluted insulators, reducing associated burdens, and improving the reliability and safety of electrical power systems. In this research, the Random Forest algorithm, one of the most prominent machine learning algorithms, was used to predict the Flashover voltage of polluted insulators. Data was cleaned and processed in a Python environment, where the algorithm demonstrated high accuracy in predicting breakdown voltage, achieving a low Root Mean Square Error (RMSE) compared to previous studies related to this field.

These findings highlight the role of artificial intelligence as an effective and reliable tool for estimating Flashover voltage and reducing the need for costly and complex laboratory experiments. This approach contributes to enhancing the reliability of electrical networks, lowering operational costs associated with maintenance and testing, and opening new horizons for applying machine learning techniques to improve the performance of electrical systems and address similar challenges.