Abstract
The article addresses the problem of improving the accuracy of assessing the technical condition of 10 kV overhead power lines, a significant proportion of which in the distribution networks of CIS countries are operated beyond their standard service life and are characterized by a high level of physical wear. Traditional diagnostic methods based on standard service life and visual inspection do not allow reliable consideration of heterogeneous and partially uncertain factors such as the age of the line, the type of poles, and the time elapsed since major overhauls were carried out. This study proposes an intelligent system for predicting the wear of 10 kV overhead power lines based on fuzzy logic methods. The developed model includes the definition of input linguistic variables, the construction of membership functions, the formation of a rule base, and the determination of the predicted wear value using a defuzzification method. To verify the effectiveness of the proposed approach, statistical data from the distribution networks of one region were used. A comparison of the model predictions with actual operational assessments demonstrated good agreement, confirmed by the calculated performance indicators mean absolute error, root mean square error, and coefficient of determination. The developed system can be used to improve diagnostic accuracy, optimize maintenance scheduling, and enhance the reliability of distribution power networks. Keywords: overhead power lines, fuzzy logic, line wear prediction, intelligent forecasting system, distribution networks.