محمد سيف الهي كرنكان

Coastal areas are important economically and socially for every country, which is why protecting them is vital. The performance and stability of seawalls are significantly impacted by wave overtopping from seawalls, which are used to protect coastal areas from wave overtopping. Furthermore, due to extreme climate changes and rising water levels, current seawalls are losing their stability; therefore, retrofit seawalls are suggested. Physical and laboratory modeling are used to predict the wave overtopping, which are converted into empirical relationships based on linear or non-linear fittings of laboratory data. A more accurate model of the wave overtopping phenomenon can be achieved with the use of intelligent computing methods which do not require knowledge of the phenomenon and are more accurate than experimental relationships; therefore, using intelligent computing methods such as artificial neural networks, a more accurate model can be obtained. In this study, a neural network is used to predict wave overtopping from retrofit seawalls. The laboratory data from Dong (2020) for four models of retrofit seawalls are used in this study. The retrofit seawalls were modeled using artificial neural networks, and the results were eva‎luated using statistical parameters. A comparison of models obtained from artificial neural networks and experimental relationships showed that the artificial neural network models had a higher accuracy. Further, each of the input parameters was entered separately into the artificial neural network to eva‎luate its effect on wave overtopping prediction. As a result, relative freeboard was the most effective parameter for predicting wave overtopping. Furthermore, to establish a relationship for predicting wave overtopping from seawalls, a decision tree was utilized. The obtained results demonstrate that the introduced statistical parameters have been enhanced, and the proposed relationship is a single equation.

ديوار ساحلي , روگذري موج , تغيير اقليم , شبكه عصبي مصنوعي

Seawall , Wave Overtopping , Climate Change , Artificial Neural Network

Mohammad Seifollahi

Dr.Abbas Yeganeh Bakhtiary