Dynamic causal weighting-based risk propagation modeling for airport movement areas

Area of Science:

• Aviation Safety
• Complex Systems Analysis
• Artificial Intelligence in Transportation

Background:

• Dynamic operational conditions significantly impact airport movement area safety.
• Understanding risk propagation is crucial for preventing incidents.
• Existing models struggle with time-varying factors and data scarcity.

Purpose of the Study:

• To develop a model for analyzing dynamic risk propagation in airport movement areas.
• To decode causal relationships between operational conditions and incidents.
• To identify critical risk factors and enable targeted control strategies.

Main Methods:

• Integration of complex network theory and causal convolutional reinforcement learning (CCRL).
• Construction of a risk propagation network with heterogeneous node classification (impedance/cumulative).
• Utilizing an enhanced grey relational model for critical risk factor identification and CCRL for dynamic propagation weight adjustment.

Main Results:

• The proposed model outperforms Dynamic Bayesian Networks (DBN) by 4.1% in prediction accuracy under data scarcity.
• Identified 63 critical risk factors influencing airport movement area safety.
• Achieved a 20% reduction in risk diffusion indices through targeted control strategies.

Conclusions:

• The novel CCRL-based model effectively deconstructs time-varying risk propagation in airport movement areas.
• The approach provides a robust method for identifying critical risk factors and mitigating their impact.
• Offers a significant advancement in aviation safety analysis and operational risk management.