Інтегрована модель оцінки та адаптивного управління експлуатаційною безпекою морського судна в умовах динамічного середовища

The article proposes an integrated model for assessing and adaptive management of the operational safety of a ship, which takes into account the temporal dynamics of risk, the technical condition of subsystems, the effectiveness of safety barriers and the impact of the external environment. The model is based on a probabilistic approach and includes generalized safety indicators: integrated risk index, safety margin, predicted time to critical condition, and system stability index. The developed algorithm provides for consistent monitoring of parameters, risk assessment in ship subsystems, modeling of barrier degradation, and dynamic aggregation of the overall risk. The verification was carried out on the basis of eight scenarios: from basic to stressful situations (power loss, combined impact, intensive maneuvering, etc.). The model features a combination of mathematical modeling and algorithmic response logic, including protection activation conditions and real-time adaptation of parameters. The use of time series and the LSTM model for risk prediction ensures proactive threat detection and decision-making before critical situations occur. The proposed system can be integrated into intelligent safety management platforms, in particular in autonomous or semi-autonomous shipping. The model has a modular structure, which ensures its flexibility and scalability for different types of vessels. Due to the use of weighting factors for the criticality of subsystems, it is possible to adapt the assessment logic to specific ship architectures and operating modes. Particular attention is paid to verifying the results based on simulated scenarios that demonstrate the dependence of the risk index and safety margin on the impact of certain factors, which allows not only post facto analysis but also the active development of protective strategies, which is relevant in the context of the transition to intelligent control systems in maritime transport. The practical value of the model lies in its ability not only to assess the current level of safety, but also to formulate recommendations for changing the speed, route, or mode of loading of the vessel. This approach ensures proactive safety management in conditions of high uncertainty and multifactorial threats