Dynamics of failure probabilities in ship power plant equipment considering cascade effects
DOI:
https://doi.org/10.33216/1998-7927-2025-295-9-5-17Keywords:
probabilistic modeling, cause-and-effect relationships, technical condition diagnostics, Bayesian networks, simulation modeling, machine learning in maintenance, equipment reliabilityAbstract
The article presents a comprehensive and scientifically substantiated approach to modeling the technical condition, degradation processes, and reliability of ship power plants (SPP) considering cascade failure effects and probabilistic dependencies between components. A hybrid diagnostic–prognostic methodology is proposed, integrating continuous-time Markov processes, Bayesian networks, gradient boosting algorithms (XGBoost), and simulation modeling within a unified analytical framework. The approach enables quantitative assessment of the dynamic evolution of reliability under complex interactions between subsystems. The interrelation between SPP components is formalized through a cascade influence coefficient matrix αᵢⱼ, which reflects how the malfunction of one unit increases the probability of failure in another. Bayesian networks are used to capture causal relationships between failures and to continuously update probabilistic assessments based on new monitoring data, while machine learning algorithms determine the most informative parameters for predictive diagnostics, such as vibration amplitude, oil temperature, and cooling system pressure. The model was trained and validated using operational data from the OREDA database and expert evaluations, demonstrating high predictive accuracy (AUC > 0.95, MAE < 4.7%). Simulation experiments identified two critical operational intervals (≈10,000 and 20,000 hours), when cascading effects lead to exponential growth of total failure probability. The cooling system and main engine were found to be the most vulnerable nodes initiating degradation chains that propagate throughout the system. The developed methodology enables integration into digital twin architectures for adaptive recalibration, anomaly detection, and risk-based maintenance optimization. The study contributes to the formation of a data-driven, cognitive basis for intelligent monitoring and predictive maintenance of maritime energy systems, enhancing their reliability, resilience, and operational efficiency under uncertainty and extended service life.
References
1. Vychuzhanin V., Vychuzhanin A. Stochastic Models and Methods for Diagnostics, Assessment, and Prediction of the Technical Condition of Complex Critical Systems. Kyiv : Liha Pres, 2025. 360 p. DOI: https://doi.org/10.36059/978-966-397-457-6
2. Vychuzhanin V., Vychuzhanin A. Intelligent Diagnostics of Ship Power Plants: Integration of Case-Based Reasoning, Probabilistic Models, and ChatGPT. A Universal Approach to Fault Diagnosis and Prognostics in Complex Technical Systems: Monograph. Lviv–Torun : Liha Pres, 2025. 412 p. DOI: https://doi.org/10.36059/978-966-397-516-0
3. Moon H., Choi J., Cha S. A multi-state Markov model to infer the latent deterioration process from the maintenance effect on reliability engineering of ships. arXiv, 2021. arXiv:2111.14368v2. DOI: https://doi.org/10.48550/arXiv.2111.14368
4. Garbatov Y., Georgiev P. Markovian maintenance planning of ship propulsion system accounting for CII and system degradation. Energies. 2024. Vol. 17, No. 16. P. 4123. DOI: https://doi.org/10.3390/en17164123
5. Morato P. G., Andriotis C. P., Papakonstantinou K. G., Rigo P. Inference and dynamic decision-making for deteriorating systems with probabilistic dependencies through Bayesian networks and deep reinforcement learning. Reliability Engineering & System Safety. 2023. Vol. 235. 109144. DOI: https://doi.org/10.1016/j.ress.2023.109144
6. Andriotis C. P., Papakonstantinou K. G., Chatzi E. N. Value of structural health information in partially observable stochastic environments. Structural Safety. 2021. Vol. 93. 102072. DOI: https://doi.org/10.1016/j.strusafe.2020.102072
7. Kamariotis A., Chatzi E. N., Straub D. A framework for quantifying the value of vibration-based structural health monitoring. Mechanical Systems and Signal Processing. 2023. Vol. 184. 109708. DOI: https://doi.org/10.1016/j.ymssp.2022.109708
8. Raptodimos Y., Lazakis I. Application of NARX neural network for predicting marine engine performance parameters. Ships and Offshore Structures. 2019. Vol. 15, No. 4. P. 412–425. DOI: https://doi.org/10.1080/17445302.2019.1661619
9. Cheliotis M., Lazakis I., Theotokatos G. Machine learning and data-driven fault detection for ship systems operations. Ocean Engineering. 2022. Vol. 216. 107968. DOI: https://doi.org/10.1016/ j.oceaneng.2020.107968
10. Zhu G., Huang L., Yin J., Gai W., Wei L. Multiple faults diagnosis for ocean-going marine diesel engines based on different neural network algorithms. Science Progress. 2023. Vol. 106, No. 4. Article 368. DOI: https://doi.org/10.1177/00368504231212765
11. Wang S., Wang J., Ding X. An intelligent fault diagnosis scheme based on PCA-BP neural network for the marine diesel engine. IOP Conference Series: Materials Science and Engineering. 2020. Vol. 782. 032079. DOI: https://doi.org/10.1088/1757-899X/782/3/032079
12. OREDA. Offshore Reliability Data Handbook. 6th ed. OREDA, 2015.
13. Chonlagarn I., Mosleh A., Modarres M. Efficient dependency computation for dynamic hybrid Bayesian network in online system health management applications. Reliability Engineering & System Safety. 2014. DOI: https://doi.org/10.36001/phmconf.2014.v6i1.2422
14. Portinale L., Codetta Raiteri D., Montani S. Supporting reliability engineers in exploiting the power of dynamic Bayesian networks. International Journal of Approximate Reasoning. 2010. Vol. 51, No. 2. P. 179–195. DOI: https://doi.org/10.1016/j.ijar.2009.05.009
15. Montani S., Portinale L., Bobbio A. Compiling dynamic fault trees into dynamic Bayesian nets for reliability analysis: the RADYBAN tool. Proceedings of the 2007 International Conference on Ubiquitous Intelligence and Computing (UIC). 2007. URL: https://ceur-ws.org/Vol-268/paper6.pdf
16. Zhang Y., Yagan O. Modeling and analysis of cascading failures in interdependent cyber-physical systems. Proceedings of the 2018 IEEE Conference on Decision and Control (CDC). 2018. DOI: https://doi.org/10.1109/CDC.2018.8618710
17. Vychuzhanin V., Vychuzhanin A. Integrated approach to creating a case-based database for diagnosing failures in ship power plants. Informatics and Mathematical Methods in Simulation. 2025. Vol. 15, No. 2. P. 155–165. DOI: https://doi.org/ 10.15276/imms.v15.no2.155
18. Vychuzhanin V., Vychuzhanin A. Integrated approach to diagnosing complex technical systems: experimental validation and multidimensional efficiency assessment. Вісник Східноукраїнського національного університету імені Володимира Даля. 2025. № 5 (291). С. 5–17. DOI: https://doi.org/10.33216/1998-7927-2025-291-5-5-17
