Publication

Abstract : Energy management is becoming one of the most important concerns in modern electrical systems for both economic and environmental factors. The capability of monitoring and predicting power consumption for energy efficiency may reduce unnecessary power waste and enhance system stability. In this paper, we develop a real-time AC power monitoring and forecasting system based on IoT, which includes an ESP8266 microcontroller, a ZMPT101B voltage sensor, and an ACS712 current sensor. The system measures the AC voltage, current, and power and energy consumption continuously and sends the data to a web server for real-time monitoring. Moreover, the LSTM neural network model is adopted to forecast future power usage patterns with the aid of collected data. The model is designed based on previously collected sensor data and further tested for its accuracy. The experimental results show a very promising accuracy range of 95−98% that indicates a new path in energy management by combining IoT and DL. This work presents how microcontroller-driven monitoring devices along with deep learning methodologies could achieve intelligent energy optimization for both household and industrial sectors.