INTELLIGENT FAULT DETECTION IN ELECTRONIC CIRCUITS USING AI FOR INDUSTRY 4.0

Abstract

Electrical faults like short circuits, open circuits, fouling, and scaling pose significant challenges in
India’s power distribution systems, leading to frequent outages and safety hazards. With over 1.4
billion people depending on electricity, power interruptions cost billions annually. The objective is to
develop an efficient machine learning-based fault detection system to accurately classify electrical
faults Traditional manual fault detection relies on periodic physical inspections by technicians who
visually or electrically check equipment for faults. This process includes using handheld meters,
visual observation, and trial-and-error testing to identify fault types and locations. It is labor-intensive,
slow, and often inaccurate due to human error and environmental conditions, leading to delayed
repairs and increased system downtime. The traditional fault detection system suffers from slow
response times and limited accuracy due to reliance on manual inspections. Human errors, lack of
real-time monitoring, and insufficient data analysis capability cause frequent misdiagnosis of faults,
resulting in prolonged outages and higher maintenance costs. The motivation behind this research is
to overcome the traditional system’s slow, error-prone fault detection by leveraging machine learning.
The proposed system uses a Support Vector Classifier (SVC) to automatically classify electrical faults
into categories: 'Short_circuit', 'Nominal', 'Open_circuit', 'Fouling', and 'Scaling'. The machine
learning model processes sensor data and identifies subtle patterns distinguishing each fault type with
high accuracy. This automation accelerates fault detection and minimizes human error, enabling
prompt maintenance actions. Additionally, SVC’s kernel functions allow it to handle complex
nonlinear relationships in the data, improving diagnostic reliability. By continuously learning from
new data, the system adapts to evolving conditions, offering a scalable and robust solution to enhance
power system stability and reduce outage durations