Neural-based monitoring system for health assessment of electric transmission lines

This paper presents a real-time monitoring system based on artificial neural network (ANN) to diagnose mechanical integrity of electric transmission lines. A sensor system composed of a receiver and a transmitter generates micro-acoustic waves along the conductor, and captures reflected waves. For key information, embedded in the captured signatures, feature extraction methods and ANN-based classifiers are considered and employed. In this monitoring system, fault detection is achieved using the reflected signatures from broken strands of transmission lines. Details of the system design, procedure and the results of performance studies in a laboratory testbed are presented. Based on laboratory experimental results, this system with both MLP (multilayer perceptron) and ART (adaptive resonance theory) classifiers show satisfactory performance in fault classification.