Machine Learning Approaches for Detection and Classification of Power Quality Disturbances

Power quality disturbances (PQDs) pose a significant challenge to the reliability, efficiency, and safety of electrical power systems. With the increasing complexity of modern power grids, traditional monitoring techniques often fall short of accurately detecting and classifying these disturbances, particularly in noisy or dynamic environments. Usually, power quality disturbances occur through non-linear loads. So, for monitoring the electrical power systems, this paper implements machine learning algorithms for detecting the power quality disturbances caused by non-linear loads like voltage sag, voltage swell, and other harmonic distortions for monitoring the power. The effectiveness of this approach is validated through extensive testing on diverse datasets like linear and non-linear loads. This paper shows the potential to greatly improve power quality monitoring by making it more reliable and adaptable.