A Digital-Twin and Machine-Learning Framework for Ventilation System Optimization for Capturing Infectious Disease...

Area of Science:

• Epidemiology
• Mechanical Engineering
• Computational Science

Background:

• The 2019 pandemic heightened interest in infectious disease modeling and simulation.
• Ventilation system design is crucial for mitigating airborne transmission of diseases via respiratory emissions (e.g., coughs, sneezes).
• Existing models require rapid computation for real-time optimization.

Purpose of the Study:

• To develop a combined Digital-Twin and Machine-Learning framework for optimizing ventilation systems.
• To determine optimal placement and flow rates of multiple ventilation units.
• To effectively sequester airborne particles from respiratory emissions.

Main Methods:

• Development of a hybrid Digital-Twin and Machine-Learning framework.
• Integration with rapidly computable respiratory emission models (Zohdi, 2020).
• Numerical simulations to ascertain optimal ventilation parameters.

Main Results:

• The framework successfully identifies optimal configurations for ventilation units.
• Demonstrated ability to sequester particles from simulated respiratory emissions.
• Provides a scalable approach for diverse ventilation scenarios.

Conclusions:

• The Digital-Twin and Machine-Learning framework offers an effective strategy for optimizing ventilation systems.
• This approach can significantly enhance indoor air quality and reduce disease transmission.
• The study provides a foundation for advanced, adaptive ventilation control systems.