AI-Enhanced Numerical Modeling for Structural Optimization of a Conceptual Large-Scale Coal MILD-oxy Combustion Boiler.
Saved in:
| Title: | AI-Enhanced Numerical Modeling for Structural Optimization of a Conceptual Large-Scale Coal MILD-oxy Combustion Boiler. |
|---|---|
| Authors: | Yu, Weizhen1 (AUTHOR), Yu, Cong2 (AUTHOR) congy@jhun.edu.cn, Wang, Feng1 (AUTHOR), Xu, Yongyi1,2 (AUTHOR), Zou, Peng1 (AUTHOR), Wu, Wei1 (AUTHOR) |
| Source: | Energies (19961073). May2026, Vol. 19 Issue 9, p2067. 31p. |
| Subject Terms: | *Multi-objective optimization, *Coal-fired boilers, *Computer simulation, *Abatement (Atmospheric chemistry), *Structural optimization, *Combustion, *Support vector machines, *Coal combustion |
| Abstract: | To advance the design of novel clean coal-fired boilers, this study integrates artificial intelligence with numerical simulations to optimize a 130 MW conceptual boiler based on Moderate or Intense Low-oxygen Dilution (MILD) and oxy-coal combustion technologies. First, mathematical models for pulverized-coal MILD-oxy combustion are validated using experimental data from a 0.58 MW pilot-scale boiler and then applied to the full-scale 130 MW boiler. An orthogonal experimental design with four factors and five levels is employed to generate 25 simulation cases, evaluating the effects of burner nozzle configuration and furnace geometry on boiler performance. Based on the simulation dataset, mutual information analysis is conducted to identify key influencing features, guiding nine additional simulations to refine samples in critical design areas. Finally, using the complete 34 simulation data, an optimal boiler structure is identified using support vector machine and multi-objective optimization algorithms. The results indicate that both the burner circumferential diameter and the O2/CO2 inlet diameter are positively correlated with nitrogen oxide (NOx) emissions, whereas the former is negatively correlated with the wall thermal non-uniformity. After optimization, the average char burnout rate increased by 1.4%, NOx emissions decreased by 4%, and wall heat non-uniformity coefficient reduced by 1.1%, demonstrating the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR] |
| Database: | Energy & Power Source |
|
Full text is not displayed to guests.
Login for full access.
|
|
| Abstract: | To advance the design of novel clean coal-fired boilers, this study integrates artificial intelligence with numerical simulations to optimize a 130 MW conceptual boiler based on Moderate or Intense Low-oxygen Dilution (MILD) and oxy-coal combustion technologies. First, mathematical models for pulverized-coal MILD-oxy combustion are validated using experimental data from a 0.58 MW pilot-scale boiler and then applied to the full-scale 130 MW boiler. An orthogonal experimental design with four factors and five levels is employed to generate 25 simulation cases, evaluating the effects of burner nozzle configuration and furnace geometry on boiler performance. Based on the simulation dataset, mutual information analysis is conducted to identify key influencing features, guiding nine additional simulations to refine samples in critical design areas. Finally, using the complete 34 simulation data, an optimal boiler structure is identified using support vector machine and multi-objective optimization algorithms. The results indicate that both the burner circumferential diameter and the O2/CO2 inlet diameter are positively correlated with nitrogen oxide (NOx) emissions, whereas the former is negatively correlated with the wall thermal non-uniformity. After optimization, the average char burnout rate increased by 1.4%, NOx emissions decreased by 4%, and wall heat non-uniformity coefficient reduced by 1.1%, demonstrating the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR] |
|---|---|
| ISSN: | 19961073 |
| DOI: | 10.3390/en19092067 |
