SP4LC: A Method for Recognizing Power Consumers in a Smart Plug

Abstract

Electrical load classification is a crucial task related to balance management in smart electrical grids. The classification algorithms and methods enable the smart system to schedule and adjust the grid load to meet the production capabilities. Fast decision-making is key to creating a responsive grid, especially when grid operators utilize renewable energy sources such as wind or solar power. This paper proposes new approach Smart Plug for Load Classification, an active load classification system to recognize the connected devices based on their load with less than 10 seconds of measurement data. Also, we propose an IoT-capable measurement device and show the collected data’s classification results with multiple methods suited for both Edge Computing and Cloud computation.