PlugGuard: A Neural Network-Based Power Quality Control System for Plug Loads

Plug loads, which encompass household appliances, significantly affect power quality (PQ) due to harmonics. Current solutions, such as building-level active filters and smart plugs, lack effective PQ control and automated load identification. To fill this gap, we present “PlugGuard,” an innovative artificial neural network (ANN) system for plug-level PQ management. PlugGuard integrates two ANNs: one for load identification and another for controlling total harmonic distortion (THD). Through rigorous optimization techniques, our approach achieves over 94% load identification accuracy and reduces THD from
$\geq$
50% to below 5%, aligning with IEEE and IEC standards. PlugGuard offers smart filtering, identifying, predicting, and compensating for plug loads' harmonics. Its integration into modern smart plugs holds promise for enhancing plug-level PQ management. Furthermore, PlugGuard's implementation meets the growing demand for plug-level solutions, contributing to market expansion and providing value to consumers and stakeholders.