Data-driven autonomous operation of VOCs removal system

Area of Science:

• Environmental Engineering
• Chemical Engineering
• Artificial Intelligence

Background:

• Controlling volatile organic compounds (VOCs) is critical in industrial settings.
• Traditional VOC removal systems rely on fixed, expert-determined operating conditions, limiting adaptability to environmental changes.
• Real-time optimization is needed for autonomous VOC removal systems to adapt to dynamic environments.

Purpose of the Study:

• To develop and implement a data-driven autonomous operation method for optimizing VOC removal systems.
• To enhance the overall performance of VOC removal systems through real-time condition adjustment.
• To demonstrate the effectiveness of an optimization framework for VOC removal applications.

Main Methods:

• Formulated an optimization problem defining decision variables (operating parameters), environmental variables, constraints, and an objective function (system performance).
• Trained a neural network using historical system data to model system performance based on operating and environmental variables.
• Solved the optimization problem in real-time to determine optimal operating conditions based on the current system environment.

Main Results:

• The proposed method successfully optimized operating conditions for a target VOC removal system.
• Achieved improved system performance without requiring manual intervention from domain experts.
• Demonstrated the feasibility of a data-driven approach for autonomous industrial system operation.

Conclusions:

• The data-driven autonomous operation method effectively enhances VOC removal system performance.
• Real-time optimization based on environmental variables is crucial for adaptive industrial processes.
• This approach offers a scalable solution for improving efficiency and reducing reliance on manual expert control in VOC abatement.