Analysis of the Influence of Waves in the Occurrence of Accidents in the Portuguese Coast Using Bayesian Belief Networks

摘要

Sea and weather conditions are the second most frequent cause of accidents in Portuguese waters accounting for 23% of the occurrences. However due to lack of information in the Portuguese maritime accident database it is difficult to assess what this cause consists specifically, i.e., fog, large wave heights or other events. In the present study significant wave height data was introduced in a Bayesian Belief Network (BBN) model in order assess the correlation between their amplitude and certain accident typologies and related consequences (human injuries or fatalities). The results of different inferences of the BBN model show that through a simple modification of an accident model the influence of weather pattern can be assessed and specific risk factors can be identified.