iradukunda3546nj/Decentralized-TinyML-Power-Load-Balancing-for-Rural-Microgrids

This study aims to Solve a real African problem unstable rural microgrids with embedded AI for real-time control. # TinyML Microgrid Simulation ## Overview This MATLAB script implements a comprehensive simulation of a TinyML-enabled decentralized residential microgrid. The system models household energy consumption, solar PV generation, and uses lightweight machine learning for short-term load forecasting to optimize energy balancing. ## Quick Start 1. Run `tinyml_microgrid_simulation.m` in MATLAB 2. The script will automatically generate synthetic data and train the TinyML model 3. Results will be displayed through multiple professional graphs showing: - System behavior (load, generation, net demand) - Prediction performance metrics - Error analysis and correlation coefficients ## Documentation For complete technical details, mathematical modeling, and in-depth result interpretation, please refer to the **project report PDF** available in this repository. ## Key Features - Realistic household load modeling with daily patterns - Solar PV generation with irradiance profiles - TinyML-based load forecasting (6-hour horizon) - Professional visualization and performance metrics - Optimized for efficient execution ## Requirements - MATLAB R2020a or later - No additional toolboxes required **Run the script and explore the results! The PDF report provides comprehensive analysis and context.**