AI Service Platform for PV & Renewable Energy Stations

Project Background:

• Developed an AI service platform for photovoltaic (PV) and renewable energy stations for State Power Investment Digital Science Co.
• Objectives: Improve intelligent monitoring, predictive maintenance, and operational efficiency.

Data Collection & Processing:

• Sources: Temperature sensors, electrical data, drone imagery.
• Preprocessing: Time-series normalization, anomaly removal.
• Feature engineering: Temporal patterns, component-level features.

Modeling & Optimization:

• Hotspot detection: Computer vision models for drone imagery.
• Predictive maintenance: ARIMA + Decision Tree ensemble for temperature forecasting.
• Model deployment and scheduling for real-time monitoring.

Workflow:

1. Collect and preprocess energy station data.
2. Train AI models for fault prediction and hotspot detection.
3. Deploy services for station monitoring and predictive alerts.

Project Outcome:

• First year: 3 stations online, revenue increase 1M RMB.
• Second year: 12 stations online, projected revenue 2.95M RMB.
• Recognized as a technological achievement in photovoltaic intelligent inspection.