Wind Forecast Analysis Dashboard

A full-stack web application that analyzes and visualizes forecasted vs actual wind power generation using real electricity system data.
The project demonstrates API data ingestion, backend processing, interactive visualization, and forecasting error analysis.

---

Live Demo

Frontend (Vercel)
https://full-stack-wind-analysis-forecastin.vercel.app/

---

Features

• Fetches real wind generation and forecast data from the Elexon BMRS API
• Processes data through a Node.js backend
• Visualizes actual vs forecast wind generation using a React dashboard
• Performs forecast error analysis using a Jupyter notebook
• Deployed using Vercel (frontend) and Render (backend)

---

Tech Stack

Frontend

• React
• Vite
• Chart.js

Backend

• Node.js
• Express
• Axios

Data Analysis

• Python
• Pandas
• Matplotlib
• Seaborn

Deployment

• Vercel
• Render

---

Project Structure

wind-forecast-dashboard │ ├── backend │ └── Node.js API for fetching and processing wind data │ ├── frontend │ └── React dashboard for visualization │ ├── analysis │ └── Jupyter notebook for forecast error analysis │ └── README.md

API Endpoint

Example request:

GET /api/forecast?horizon=4

Example response:

[
  {
    "time": "2024-01-01T04:00:00Z",
    "actual": 6900,
    "forecast": 12000
  }
]
Forecast Analysis

Forecast error is computed as:

Forecast Error = Forecast Generation − Actual Generation

The analysis notebook evaluates:

Mean forecast error

Error distribution

Forecast bias patterns

Results indicate that wind forecasts often overestimate actual generation, highlighting systematic forecasting bias.

Local Setup
Backend
cd backend
npm install
node server.js
Frontend
cd frontend
npm install
npm run dev

Frontend runs at:

http://localhost:5173

Environment Variable

Create .env in the frontend folder:

VITE_API_URL=http://localhost:5000

Data Source

Elexon BMRS Electricity Data API
https://data.elexon.co.uk/bmrs/api/v1

Author

Sneha Ilager
Computer Science Engineering Student