| The shipping industry is a significant support for international trade and marine resources exploitation.In recent years,along with the development of the shipping industry,serious maritime accidents have also occurred,and the safety level of waterway transportation has not reached a satisfactory level.Human factors are generally considered to be the direct or indirect causes of 80% of marine traffic accidents.It is of great theoretic significance and application value to explore the human errors and human performance of ocean-going vessels’ deck officers,so as to improve the shipping industry’s safety and reduce maritime accidents.This study identifies common human factors in marine accidents,and constructs risk scenarios to enable the anslysis of deck officer’s emotion and mental workload within experimental scenarios.Then,it combines the ship bridge simulator with Electroencephalogy(EEG)and functional Near-Infrared Spectroscopy(f NIRS),technologies to conduct the experimental study on psychological factors of deck officers.Using risk assessment and analysis methods quantifies deck officers’ emotion and mental workload.Morever,recognition models of emotion and mental workload are constructed to study its relevance to human errors and decision-making behaviour of deck officers.The main content and innovation of the thesis are summarised as follows:1.The model of human factors in maritime transportation is established based on the data mining from maritime accident reports.Then data-driven Bayesian network modeling is conducted using Tree-augmented Naive Bayes.It illustrates the impact of different risk factors on different types of maritime accidents.The sensitivity analysis is carried out,and accident types influenced by multiple factors are identified by scenario analysis.At the same time,a high-risk accident scenario is developed for subsequent experimental research.2.The methods of deck officers’ emotion calibration and quantification are studied.Based on the experiment on the ship bridge simulator platform,EEG,questionannaire and behavioural data of deck officers are collected to quatify deck officer’s emotion.The deck officer’s emotion state during watchkeeping influences their attention,judgment and decision making,so as to affect their behaviours in the study.Wavelet analysis is used to extract EEG data’s features.Then they are trained as support vectors in a classifier to identify emotion types.In this way,the emotion of ship ofiicers are quantified using ship bridge simulation with average accuracy rate of 77.55%.3.The changes in deck officers’ oxygenation and mental workload are studied.Based on the experiment on ship bridge simulator platform,f NIRS,questionnaire and behaviour data of deck officers are collected to quantitatively analyse deck officers’ oxygenation and mental workload with different nautical experiences under different task difficulties.The study proves that experienced deck officers make earlier decisions to avoid collision in ship encounter situations.Right dorsolateral prefrontal cortex of deck officers shows increased oxygenation during decision making,while it reduces at the end of the watchkeeping.The experienced group has higher oxygenation during watchkeeping.The quantitative analysis of the deck officer’s mental workload explains cognitive demand and better performance in decision making.4.The correlation between deck officer’s emotion and human errors,and the association between mental workload and human performance are explored.EEGbased quantitative analysis of deck officers’ emotion show that negative emotion is more likely to cause human errors.Less negative emotion is the most dangerous emotional state that causes human errors during navigation,compared to extreme negative emotion.Quantitative analysis of deck officers’ mental workload using f NIRS demonstrates that the deck officer becomes more efficient in the decision making as the density coefficient of functional brain connectivity decreases.And the better behaviour is relevant to increased activity of the right dorsolateral prefrontal cortex,decreased density connection,and increased clustering.The thesis integrates knowledge of psychology and neuroscience with transportation engineering.The results help the shipping industry and maritime authorities to deeply understand deck officers’ cognitive load,provide theoretical guidance for risk management of deck officers’ errors and behaviours,and offer evaluation critiria for crew training and performance quantification. |