🌬️ WindPulse

Heat Map-Based Clustered Diurnal Wind Power Density Analysis for Optimized Maintenance Scheduling of Wind Farms in India

WindPulse is an advanced analytics platform that leverages heat map visualizations and clustering algorithms on diurnal wind patterns to enable data-driven maintenance planning for Indian wind farms — maximizing uptime and energy production efficiency.

📊 Key Results

Max power loss reduction via optimized scheduling

Reliability with risk-controlled approach

Publication-ready visualizations

Indian wind farm regions analyzed

✨ Key Features

📈 Diurnal Wind Pattern Analysis

Comprehensive analysis of hourly wind power density patterns throughout 24-hour cycles across all seasons.

🗺️ Heat Map Visualization

Intuitive color-coded contour representations of wind power density across time and location.

🤖 ML Clustering

PCA, t-SNE, and UMAP-based clustering to identify similar wind patterns across different wind farm sites.

⚙️ Optimized Scheduling

Risk-controlled maintenance window recommendations based on low wind power density periods.

🇮🇳 India-Specific Data

Tailored for Indian wind farm locations across Maharashtra, Tamil Nadu, Gujarat, Karnataka, Rajasthan, and Andhra Pradesh.

🔮 Predictive Analytics

Weibull distribution and Markov model-based forecasting for proactive maintenance planning.

🎯 Problem Statement

Wind farms require regular maintenance to ensure optimal performance and longevity. Scheduling maintenance during high wind power periods results in significant energy production losses. WindPulse addresses this by: 1. **Analyzing historical wind power density data** — process years of wind data to identify patterns 2. **Identifying consistent low-power periods** — find optimal maintenance windows across diurnal cycles 3. **Clustering wind farms with similar patterns** — group sites for coordinated maintenance 4. **Recommending optimal maintenance windows** — minimize production impact with data-driven scheduling

📸 Gallery Preview

Browse through 60 high-quality visualizations including heat maps, contour plots, clustering diagrams, Sankey flows, and more.

🔍

Wind Power Density Contour

🔍

Sankey Flow Diagram

🔍

PCA Clustering 2D

View Complete Gallery (60 images) →

🚀 Quick Start

# Clone the repository
git clone https://github.com/Samsomyajit/windpulse.git
cd windpulse

# Install dependencies
pip install -r requirements.txt

🛠️ Technology Stack

| Layer | Tools | |-------|-------| | **Data Analysis** | Python, NumPy, Pandas, SciPy | | **Machine Learning** | Scikit-learn, tslearn | | **Visualization** | Matplotlib, Seaborn, Plotly | | **Statistical Methods** | Weibull distribution, Markov models, Bootstrap |

📄 License

This project is licensed under the Apache License 2.0.
For questions or collaboration, please open an issue on GitHub.