Experimental investigation on wave transmission, reflection and dissipation characteristics of rubble mound breakwater models are time consuming and expensive. However, such studies are required for designing the rubble mound breakwaters for marine structures in an optimal condition. In order to overcome such problems many researchers used various soft computing techniques such as Artificial Neural Network (ANN), Adaptive Neuro Fuzzy Interference System (ANFIS), Genetic Programming (GP), Support Vector Machine (SVM) etc, in order to predict the design factors in the field of coastal engineering. The current work proposes Genetic Programming (GP) as a modeling tool to evolve mathematical models for the behavior of single and double breakwaters. Based on the detailed experimental data, GP models were performed to predict the reflected wave height (Hr), wave height on the breakwater (H5) and transmitted wave height (Ht) by considering with and without trigonometric effects of those breakwaters. The quality of predictability of the present model is measured by the statistical parameter, RMSE (Root Mean Square Error). Since the waves were more complex in nature, it is very essential in considering the trigonometric function’s effect in the modeling aspects. It is evident that, the GP model accurately described the non linear complex effects.

中文翻译：

---

使用遗传编程工具对单、双碎石堤防波堤行为进行建模研究

碎石堤防波堤模型的波传输、反射和耗散特性的实验研究既费时又费钱。然而，这些研究需要在最佳条件下设计用于海洋结构的碎石堤防波堤。为了克服这些问题，许多研究人员使用了各种软计算技术，如人工神经网络 (ANN)、自适应神经模糊干扰系统 (ANFIS)、遗传编程 (GP)、支持向量机 (SVM) 等，以预测海岸工程领域的设计因素。目前的工作建议将遗传编程 (GP) 作为建模工具，为单防波堤和双防波堤的行为演化数学模型。根据详细的实验数据，通过考虑有和没有这些防波堤的三角效应，执行 GP 模型来预测反射波高 (Hr)、防波堤上的波高 (H5) 和透射波高 (Ht)。本模型的可预测性质量通过统计参数 RMSE（均方根误差）来衡量。由于波浪在本质上更为复杂，因此在建模方面考虑三角函数的影响是非常必要的。显然，GP 模型准确地描述了非线性复杂效应。由于波浪在本质上更为复杂，因此在建模方面考虑三角函数的影响是非常必要的。显然，GP 模型准确地描述了非线性复杂效应。由于波浪在本质上更为复杂，因此在建模方面考虑三角函数的影响是非常必要的。显然，GP 模型准确地描述了非线性复杂效应。