Abstract

The application of Artificial Intelligence (AI) techniques has become popular in science and engineering applications since the middle of the twentieth century. In this present study, three AI techniques (ANFIS, GP and ANN) have been used for forecasting streamflow into Shakkar watershed (Narmada Basin), India. The models have been used considering previous streamflow and cyclic terms in the input vector to provide a suitable time series model for streamflow forecasting. To evaluate the model performance, RMSE, MAE, CORR and CE were employed. Results showed that the ANFIS has the best performance in forecasting streamflow time series for Shakkar watershed. The GP and ANN are in the 2nd and 3rd ranks, respectively. According to the results, in all the AI methods (ANFIS, GP and ANN), the model with cyclic terms had better performance compared to those models not considering periodic nature and being applied by only considering the previous streamflow.

Graphical Abstract

中文翻译：

---

基于人工智能技术的水流预测

摘要

自 20 世纪中叶以来，人工智能 (AI) 技术的应用在科学和工程应用中变得流行。在本研究中，三种人工智能技术（ANFIS、GP 和 ANN）已被用于预测流入印度 Shakkar 流域（纳尔马达盆地）的流量。这些模型已用于考虑输入向量中先前的流量和循环项，为流量预测提供合适的时间序列模型。为了评估模型性能，采用了 RMSE、MAE、CORR 和 CE。结果表明，ANFIS在预测Shakkar流域水流时间序列方面的性能最好。GP 和 ANN 分别位于第 2 和第 3 行。根据结果​​，在所有的 AI 方法（ANFIS、GP 和 ANN）中，

图形概要