Big Data Intelligence for Stealthy Anomaly Detection and Mitigation in Substations to Secure Smart Grid.

Description

The dataset captures real-time operations of the key power systems monitoring devices in a medium voltage (20 kV) outdoor substation located in a remote area. • The datasets provide continuous monitoring of three-phase current and voltage signals, power quality indicators, and circuit breaker status using IEC 61850 (GOOSE and sampled measured values) and IEC 104/ IEEE 802.3 communication protocols. • The datasets are highly valuable for designing, developing, and testing AI-based hybrid models and algorithms (e.g., the HML) for stealthy anomaly detection, classification, and predictive maintenance in substations and smart grids. • The time series data represents substation behaviour under realistic normal and anomalous scenarios, and could enable utilities to evaluate anomaly detection systems, intrusion detection mechanisms, and potential defence responses in substations to safeguard smart grids. • Various stakeholders, including grid operators, researchers, cybersecurity analysts, SCADA developers, policymakers, and planners, can benefit from the datasets to evaluate stealthy vulnerabilities, network resilience, and to design digital twin and cyber-physical defence strategies in SGs.

Steps to reproduce

Set Up the Project Environment 1. Install VS Code and Required Extensions e.g.: Python Jupyter GitLens (for version control) Docker (if using containers) ------------------------------------------------------------------ 2. Clone or Create the Project From the same directory where you stored the data, such as: C:\Users\---\---\Desktop\Data In Brief\Code and Data\HML\ ------------------------------------------------------------------ 3. Set Up the Virtual Environment (Optional but Recommended) In the terminal (inside VS Code): bash Copy Edit # For Python python -m venv venv source venv/bin/activate # On Windows: venv\Scripts\activate # Install dependencies pip install -r requirements.txt ----------------------------------------------------------------- 4. Load or Access the Dataset Place the dataset in the correct directory (if not already there) If it's downloaded from the internet or a database, make sure the code includes that functionality Check if .env or config files are used for paths, credentials, etc. ----------------------------------------------------------------- 5. Open the Notebook or Script File For Jupyter notebooks: Open .ipynb files directly in VS Code. For Python scripts: Open .py files. Make sure to read through the comments and understand the structure: Data loading Preprocessing Analysis/modeling Output/export ----------------------------------------------------------------- 6. Run the Code Option A: Jupyter Notebook Run cells one by one or use Run All (top menu) Option B: Python Script Press F5 to run the script Or use: bash Copy Edit python script_name.py ---------------------------------------------------------------- 7. Inspect the Output Check if the output files (CSV, images, logs, models) are generated Validate that the reproduced data matches expected outputs ---------------------------------------------------------------- 8. Troubleshoot as Needed Common issues: Missing dependencies → Check requirements.txt File paths → Use absolute/relative paths carefully Data access issues → Check environment variables, API keys, permissions ----------------------------------------------------------------

Institutions

• Bahauddin Zakariya University

  Punjab, Multan
• Teknologian tutkimuskeskus VTT Oy

  Espoo
• Al-Ahliyya Amman University

  Amman

Categories

Funders

• VTT Technical Research Centre of Finland

  Uusimaa, Finland

Licence