With the rapid development of the global shipping industry, maritime traffic continues to grow, and maritime traffic risks are becoming increasingly prominent, posing serious threats to economic development, the ecological environment, and public safety. In this context, this study develops an environmental data-driven dynamic Bayesian network (DBN) model to simulate the dynamic evolution process of maritime traffic risks from the massive data of complex shipping systems. Firstly, based on the systems theoretic accident model and processes (STAMP) accident causation analysis framework, risk influencing factors (RIFs) are identified through the analysis of maritime accident report systems. Secondly, addressing the dynamic nature of maritime risks, a novel transition probability matrix (TPM) learning mechanism integrating environmental data is proposed, constructing a DBN model capable of characterizing temporal risk performance. Finally, a case study of typical routes along the Chinese coast reveals that risk performance evolution exhibits significant spatiotemporal heterogeneity across sea areas, with the East Sea and South China Sea regions being the most prominent. Their fluctuations are highly correlated with seasonal meteorological and hydrological changes, and the distribution of accident risks is also closely associated with extreme weather events such as typhoons. Sensitivity analysis validates the model's reliability. This study provides a quantitative tool for the dynamic risk management of intelligent shipping systems and offers policy insights for intelligent maritime transportation safety regulation.

This study develops a novel deep learning framework for bank distress prediction, designed to overcome the limitations of static network analysis and to enhance model interpretability. We propose a Spatio-Temporal Attention Network that uniquely captures the time-varying nature of systemic risk. Methodologically, it introduces two key innovations: (1) a dynamic interbank network whose connection weights are adjusted by the volatility of the Shanghai Interbank Offered Rate (SHIBOR), reflecting real-time market liquidity changes; and (2) a dual spatio-temporal attention mechanism that identifies critical time steps and pivotal contagion pathways leading to a distress event. Empirical results demonstrate that the model significantly outperforms traditional benchmarks across key metrics including accuracy and F1-score. Most critically, the architecture proves exceptionally effective at reducing Type II errors, substantially minimizing the failure to identify at-risk banks. The model also offers high interpretability, with attention weights visualizing intuitive risk evolution patterns. We conclude that incorporating dynamic, liquidity-adjusted networks is crucial for superior predictive performance in systemic risk modeling.