MEO37th/power-grid-watch

predicting and simulating TCN National Grid using Nigeria's 24-bus 330KV actual Data # GridWatch - Real-time Grid Stability Monitor A complete, modular solution for monitoring power grid stability using machine learning models trained on 24-bus and 50-bus systems. ## Features - Real-time stability prediction with visual gauge - Voice alarm system for critical events - Demo replay mode with historical fault scenarios - Support for 24-bus and 50-bus system models - Risk trend visualization - Mock mode for immediate testing without trained models ## Prerequisites - Node.js (v16+) - Python (3.9+) - pip ## Quick Start ### Backend Setup ```bash cd backend pip install -r requirements.txt uvicorn api:app --reload ``` The API will run on localhost ### Frontend Setup ```bash cd frontend npm install npm run dev ``` The app will run on localhost ## Project Structure ``` grid-guardian-ai/ ├── backend/ │ ├── api.py # FastAPI server │ ├── model_loader.py # Model management │ ├── adapter.py # Input transformation │ └── requirements.txt ├── frontend/ │ ├── src/ │ │ ├── App.jsx # Main application │ │ ├── components/ │ │ │ └── Gauge.jsx # Risk gauge component │ │ ├── main.jsx │ │ └── index.css │ ├── index.html │ ├── package.json │ └── tailwind.config.js ├── models/ # Place your .pkl files here └── assets/ ``` ## Adding Your Models Place your trained model files in the `models/` directory: - `24bus_model.pkl` for 24-bus system - `50bus_model.pkl` for 50-bus system If models are not found, the system will use mock predictions for demo purposes. ## Demo Script (60 seconds) 1. (0:00) "This is GridWatch, a real-time stability monitor based on thesis work with 24 and 50-bus systems." 2. (0:15) Click '24-Bus', move sliders - "System in stable state. Low risk, green indicators." 3. (0:25) Click 'Replay Demo' - "Replaying historical fault scenario..." 4. (0:40) Gauge hits Red, Voice triggers - "AI detects voltage collapse precursor 3 seconds befo …