This project aims to develop an advanced intrusion detection system (IDS) for string inverters in solar systems. By leveraging machine learning and cloud technologies, the IDS can detect anomalies and potential cyber-attacks on string inverters, ensuring the safety and efficiency of solar energy systems.
- Feature Extraction: Extracts important features from raw data collected from string inverters, such as frequency of communication between devices, type of data transmitted, and time of day of transmission.
- Machine Learning Model: Uses transfer learning to train on large datasets of network traffic data, identifying normal behavior patterns and detecting anomalies.
- Real-time Monitoring: Provides real-time alerts about potential threats and system status.
- Scalability: Designed to handle large amounts of data and can be deployed in various environments, from small-scale PV systems to large-scale energy infrastructure.
- Google Cloud Platform (GCP): The primary platform for development, deployment, data storage, and machine learning.
- Python: The main programming language used for algorithm development and data processing.
- TensorFlow & scikit-learn: Machine learning libraries for model development and training.
- Matplotlib & Seaborn: Data visualization tools for insights and analysis.
- Jupyter Notebook: Used for data exploration, algorithm validation, and cross-validation.