Reduced-order modeling of multiphase flow in fluidized beds using proper orthogonal decomposition and neural ODEs

Key Points

• Reduced-order modeling demonstrates improved accuracy of multiphase flow simulations in fluidized beds under various conditions.
• Findings indicate a 30% reduction in computational resources while maintaining simulation fidelity using neural ODEs.
• Analysis utilizes proper orthogonal decomposition for effectively capturing fluid dynamics in complex multiphase interactions.
• Improvements highlight the potential for broader application in industrial multiphase flow systems, benefiting efficiency and cost.