FAULT DETECTION SYSTEM IN ANCHORING EYE BOLT.
Anchoring eye bolts, electromagnetic parameters, fault detection, preventive maintenance, random forest algorithm, transmission lines.
Eyebolts are components responsible for connecting insulator chains to poles and concrete beams, supporting the load of overhead electrical wiring. Hidden structural failures in eyebolts can lead to their rupture and, consequently, cause serious accidents and interruptions in the power supply, exposing operators to penalties and financial fines. This thesis presents a system for detecting structural failures in anchoring eyebolts, based on the analysis of the S11 magnitude through a random forest classifier trained with high-fidelity measurements and simulated signals. The proposed methodology is completely non-invasive and does not require dismantling the electrical infrastructure. The high accuracy of the presented results suggests that the proposed method could improve the efficiency of preventive maintenance routines carried out on eyebolts and, consequently, increase the reliability of power distribution systems.