Use of Artificial Neural Networks to Enhance Container Port Safety Analysis Under Uncertainty

Hani Yami, Ramin Riahi, Jin Wang and Zaili Yang ()
Additional contact information
Hani Yami: King Abdulaziz University
Ramin Riahi: Columbia Shipping Management (Deutschland) GmbH
Jin Wang: Liverpool John Moores University
Zaili Yang: Liverpool John Moores University

A chapter in Advances in Reliability and Maintainability Methods and Engineering Applications, 2023, pp 265-291 from Springer

Abstract: Abstract This chapter proposes a modified failure mode effect analysis (FMEA) approach using Artificial Neural Networks (ANNs) to evaluate and predict the operational risks of container terminals. It effectively integrates two established methods in one framework to realise complex risk analysis from a whole system perspective, including fuzzy rule based Bayesian networks (FRBN) for risk analysis of particular hazards in ports and fuzzy evidential reasoning (FER) for safety evaluation of ports in a systematic way. During this process, ANNs are integrated with FRBN and FER respectively to create two sub-models. The first sub-model is FRBN-ANN that incorporates Bayesian networks (BNs) with ANNs to facilitate risk prediction of each identified hazard in a container port. The second sub-model is FER-ANN, which uses ANNs to simulate the FER method to ease the aggregation of all the hazards to obtain the safety level of the port. Finally, the two sub-models are combined into a single safety model, which can help simplify risk prediction, and realise real-time safety evaluation of ports at hazard or whole system levels. The Levenberg–Marquardt (trainlm) back-propagation algorithm trial and error approach was used to determine the optimal ANN architecture. The proposed ANN model produced small deviations that indicate high predictive accuracy with satisfactory determination coefficients (i.e., the regression) for forecasting operational risks of container ports. It provides an effective risk prediction tool for complex port safety systems, and significantly simplifies the port safety analysis and prediction in a feasible, versatile, and accurate manner. It, through the black box approach of ANN, provides a mathematically unsophisticated solution and hence aids the visualisation of risk analysis outcomes without the need of the end users to understand the complicated computing process of the risk inference. It makes significant contributions to port safety analysis and management in practice.

Date: 2023
References: Add references at CitEc
Citations:

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:ssrchp:978-3-031-28859-3_11

Ordering information: This item can be ordered from
http://www.springer.com/9783031288593

DOI: 10.1007/978-3-031-28859-3_11

Access Statistics for this chapter

More chapters in Springer Series in Reliability Engineering from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().