A rainwater control optimization design approach for airports based on a self-organizing feature map neural network...

Area of Science:

• Environmental Engineering
• Artificial Intelligence
• Urban Hydrology

Background:

• High overflow rates in pipe network inspection wells and low drainage efficiency are significant challenges in urban water management.
• These issues often stem from imprecise parameter design in rainwater control measures, such as pipe diameter and green infrastructure ratios.
• The Daxing District International Airport in Beijing faces these challenges, necessitating optimized rainwater management.

Purpose of the Study:

• To propose a novel rainwater control optimization design approach using a self-organizing feature map neural network (SOFM) model.
• To address low precision parameter design issues in rainwater control measures like pipe network diameter and green roof area ratio.
• To enhance the control of airport rainfall and promote sustainable drainage systems.

Main Methods:

• Development and application of a self-organizing feature map neural network (SOFM) model for optimization.
• Optimization of rainwater pipe network parameters, including diameter and slope.
• Integration and adjustment of green infrastructure (GI) parameters, such as sinking green area and green roof area.

Main Results:

• The SOFM model demonstrated significant improvements in managing a hundred-year torrential rainstorm.
• Overflow rates in pipe network inspection wells were reduced by 36% to 67.5%.
• Drainage efficiency increased by 26.3% to 61.7% compared to initial design parameters.

Conclusions:

• The proposed SOFM-based optimization design effectively addresses challenges in pipe network overflow and drainage efficiency.
• The approach facilitates reasonable control of airport rainwater and contributes to building sponge airports and sustainable drainage systems.
• Optimized design parameters, including pipe network and green infrastructure elements, are crucial for resilient urban water management.