Primary supervisor

Isma Farah Siddiqui

This project aims to develop a modular and distributed digital twin framework focused on simulating, monitoring, and analysing smart grid data. The framework will represent virtual models of grid components (e.g., loads, meters, nodes) and synchronise them with real-time or simulated data streams using distributed systems principles.

The testbed will support:

  • Digital twin modeling of grid entities (smart meters).
  • Distributed data ingestion and synchronization using messaging protocols.
  • Time-series data storage for historical and real-time analysis.
  • Interactive dashboards for monitoring grid behaviour.
  • Basic analytics for load profiling and anomaly detection.

This foundational framework will be extensible for future integration with machine learning models, renewable energy systems, and real-world IoT data sources.

Student cohort

Single Semester
Double Semester

Aim/outline

Project Outline:

The project will need working on few essential and optional modules, discussed below:

  1. Digital Twin Modelling
    This module defines the virtual representations of grid entities. Each twin is modeled using structured data formats such as JSON or semantic schemas (optional). These models capture the attributes and relationships of grid components, forming the foundation of the digital twin layer.

  2. Distributed Data Ingestion
    To simulate real-time behaviour, this module streams data from multiple sources, either simulated or real, into the system. Technologies like Apache Kafka or MQTT are used to handle distributed messaging, ensuring that data flows efficiently and reliably to the appropriate digital twin instances.

  3. Time-Series Data Storage
    All incoming data is stored in a time-series database to support historical analysis and real-time querying. Tools like InfluxDB or TimescaleDB are ideal for this purpose, offering high-performance storage and retrieval of grid metrics such as voltage, current, and power consumption.

  4. Twin State Synchronisation
    This module ensures that the state of each digital twin remains consistent and up to date across distributed nodes. It uses coordination tools like Kafka (for event streaming) and Redis or Zookeeper (for caching and synchronisation) to manage state updates and system-wide consistency.

  5. Monitoring and Visualisation
    A user-facing dashboard is developed to visualise the real-time behaviour of the grid and its digital twins. Tools like Grafana or Streamlit are used to create interactive visualisations, allowing users to monitor key metrics, detect anomalies, and explore historical trends.

  6. Analytics Module
    This module performs lightweight analytics on the collected data. It includes basic functions such as load profiling, anomaly detection, and cost estimation using Python libraries like Pandas, SciPy, and Matplotlib. These insights help evaluate the performance and behaviour of the grid under different conditions.

URLs/references

https://arxiv.org/pdf/2104.07904

Required knowledge

The required knowledge includes, but is not limited to, the following areas:

  • Basics of smart grid systems
  • Understanding of digital twin concepts
  • Data collection and monitoring techniques
  • Time-series data analysis
  • Fundamentals of machine learning
  • Basics of distributed computing
  • Concepts of cloud and edge computing
  • Programming skills (e.g., Python, Java)
  • Familiarity with IoT (Internet of Things) and SCADA (Supervisory Control and Data Acquisition) systems