Increasing of wave simulation accuracy through generating of local wind with artificial neural network technique in Bushehr shallow water

The waves affect the vertical movements of the vessels, thereby endangering the safety of their movement in coastal waters. This study has been conducted to accurate wave prediction in the shipping channel of Bushehr port in order to ensure safe and efficient maritime transportation of this region. First, the wave simulation was performed using PMODynamicsI software and using GFS predictable wind field. By applying the optimal settings in the model and comparing the model results with the wave measured values at the shipping channel location, the accuracy of the verification results was equal to 70%. The results showed that the use of large-scale and global wind caused a decrease in modeled wave height and lack of proper accuracy in wave direction simulation during the occurrence of storms with a sudden change of direction, which due to the importance of maritime safety discussion, accurately prediction of the wave by using predictable local wind has been considered. This has been done through artificial neural network technique and design of different scenarios, which has generated wind with local precision by training the GFS global wind field through meteorological wind data. This wind has been used in Bushehr wave model, which has led to a 20% and 15% increase in accuracy for the direction and height of the modeled wave, and a 10% decrease in the root mean square error.