Welcome to F97.BE!

Python, Photons, and Predictions.

Solar

Solar Power Forecast

At the heart of my forecasting work now lies the groundbreaking Fusion forecast, a key step in the evolution of solar prediction: it unites the learning power of AI with the physics of GTI, dynamically learning the optimal balance between them, applying adaptive weights and offsets, and delivering more robust results under unusual weather conditions. Alongside Fusion, you can explore my 24-hour energy production forecast for the House Roof System, generated with a custom AI model trained on more than 1500 days of historical weather and solar production data. The dashboard visualizes real-time AI predictions, GTI-based forecasts, and actual inverter production in a single chart, and includes key performance metrics (R², MAE) as well as historical accuracy tables.

Solar Power Monitoring

While our system began with simple dashboards, it has evolved into a powerful, custom-built energy management platform. At its core is a suite of lightweight Python programs designed specifically for solar forecasting, data processing, and system monitoring. These tools provide enhanced real-time and historical insights, deeper analytics, and fine-grained performance tracking across both solar arrays.

Data from the inverters is fetched directly via APIs and processed locally on a Raspberry Pi. From there, the system prepares structured datasets and uploads them to a public-facing web server, where they are rendered in clear, responsive HTML dashboards. This approach ensures full control over the data pipeline while maintaining maximum flexibility and transparency.

All solar production and energy usage data is continuously stored in a local PostgreSQL database, providing a robust, scalable foundation for long-term retention and advanced analytics. With this architecture, the system can generate detailed daily, monthly, and yearly statistics, perform performance comparisons across seasons, detect anomalies, and optimize charging and consumption behavior.

Ultimately, this self-contained, end-to-end solution gives me real-time awareness and long-term perspective — empowering data-driven energy decisions and sustainable smart home management.

Click below to dive into the detailed data for each system over the last 24 hours and see how much energy we’re generating right now!

House Roof System

  • 24 × 370W solar panels
  • South-facing (180°) for optimal energy generation during peak sunlight hours
  • Designed to capture the most sunlight throughout the day, ensuring consistent power production for our home

Studio Roof System

  • 4 × 410W solar panels
  • West-facing (270°) to extend solar production into the evening
  • Strategically placed to capture the late-afternoon and evening sun, complementing the house system’s output

Compare energy production of both our systems.

Berechne, wie viel Leistung deine Solaranlage bei Sommerhitze verliert – mit realen Wetterdaten, Sonnenstunden und typischen Modulen: Zum Leistungsverlust-Rechner.

home | top