Abstract
Energy management has become a critical concern in modern society due to the increasing demand for electricity and the need for efficient utilization of resources. Traditional energy monitoring systems often lack real-time analysis, predictive capability, and intelligent decision-making features. To address these limitations, this project presents a Real-Time Power Monitoring and Prediction System using Machine Learning.The proposed system integrates real-time data acquisition, hybrid power calculation, machine learning prediction, and web-based visualization into a unified framework. Electrical parameters such as voltage, current, and power are collected through an API and processed using data cleaning and preprocessing techniques. A hybrid approach is employed to enhance accuracy by combining API-provided power values with calculated power (V × I).
The system utilizes a Random Forest machine learning model to predict total power consumption based on input features. The predicted values are further classified into categories such as low, medium, and high to improve interpretability. The processed data is stored in an SQLite database and displayed through a Flask-based web application, enabling users to monitor device-wise consumption, total power usage, and estimated electricity bills in real time.Graphical visualizations are incorporated to analyze power trends, device-level consumption, and cost relationships, providing meaningful insights for energy optimization. The results demonstrate that the system achieves accurate prediction, efficient monitoring, and improved decision-making capability.Overall, the proposed system offers a scalable, cost-effective, and intelligent solution for smart energy management, with potential applications in smart homes, industrial systems, and IoT-based environments.
Keywords
Real-Time Power Monitoring, Machine Learning, Random Forest, Energy Consumption, Hybrid Calculation, Flask Web Application, SQLite Database, Power Prediction, Electricity Billing, Data Visualization.
Introduction
The rapid growth in electricity consumption and the increasing complexity of modern energy systems have created a strong need for efficient monitoring and management solutions. Traditional energy monitoring methods are often limited to manual readings or static systems that do not provide real-time insights or predictive capabilities. As a result, users lack awareness of their energy usage patterns, leading to inefficient consumption and increased electricity costs.
With the advancement of digital technologies, the integration of real-time data processing, web applications, and machine learning has opened new possibilities for intelligent energy management systems. Real-time monitoring allows continuous tracking of electrical parameters such as voltage, current, and power, enabling users to make informed decisions. However, monitoring alone is not sufficient; predictive analysis and classification are essential to fully understand consumption behavior and optimize energy usage.
This project proposes a Real-Time Power Monitoring and Prediction System using Machine Learning, which combines data acquisition, processing, prediction, and visualization into a single platform. The system collects electrical data from an API source and processes it through data cleaning and feature extraction techniques. Power is calculated using the fundamental electrical relation (V × I), and a hybrid approach is applied by combining calculated and API-provided values to improve accuracy.
To enhance the intelligence of the system, a Random Forest machine learning model is implemented for predicting total power consumption based on input features such as voltage and current. The predicted values are further categorized into different levels (low, medium, high), allowing users to easily interpret energy usage patterns. This classification helps in identifying high consumption scenarios and supports efficient energy management.
The system is implemented using a Flask-based web application, which provides an interactive interface for displaying real-time data, device-wise consumption, total power, and estimated electricity bills. All processed data is stored in an SQLite database, ensuring efficient data management and retrieval. Additionally, graphical visualizations such as power trends, device comparisons, and cost analysis are used to improve understanding and support decision-making.
Overall, the proposed system addresses the limitations of traditional energy monitoring methods by providing real-time insights, predictive analysis, and user-friendly visualization. It offers a scalable and efficient solution for modern energy management, with potential applications in smart homes, industrial monitoring systems, and IoT-based environments.


Objectives
• To develop a system for real-time monitoring of electrical parameters such as voltage, current, and power
• To implement accurate power calculation using the formula (V × I)
• To enhance accuracy using a hybrid calculation method by combining API data and computed values
• To design and implement a Random Forest machine learning model for predicting total power consumption
• To classify power consumption into categories such as Low, Medium, and High for better interpretation
• To store collected and processed data efficiently using an SQLite database
• To develop a Flask-based web application for visualizing real-time and historical data
• To generate and display graphical representations of power consumption and billing
• To calculate the electricity bill based on power usage
• To provide an intelligent system that supports efficient energy management and decision-making

Block Diagram