Accurate Discharge Coefficient Prediction of Streamlined Weirs by Coupling Linear Regression and Deep Convolutional Gated Recurrent Unit

EasyChair Preprint 7761

Abstract

Streamlined weirs, as a nature-inspired type of weirs, have gained tremendous attention among hydraulic engineers mainly due to their well-known performance with high discharge coefficient. Computational fluid dynamic (CDF) is considered as a robust tool to predict discharge coefficient. To bypass the computational cost of CFD-based assessment, the present study proposes data-driven modelling techniques, as an alternative to CFD simulation, to predict discharge coefficient based on an experimental dataset. To this end, after splitting the dataset by k-fold cross-validation technique, the performance assessment of classical and hybrid machine-deep learning (ML-DL) algorithms is undertaken. Amongst ML techniques, linear regression (LR), random forest (RF), support vector machine (SVM), k-nearest neighbours (KNN), and decision tree (DT) algorithms are studied . In the context of DL, long short-term memory (LSTM), convolutional neural network (CNN), gated recurrent unit (GRU) and their hybrid forms such as LSTM-GRU, CNN-LSTM and CNN-GRU techniques are compared by different error metrics. It is found that the proposed three-layer hierarchical DL algorithm consisting of a convolutional layer coupled with two subsequent GRU levels, which is also hybridized by LR method (i.e., LR-CGRU), leads to lower error metrics. This paper paves the way for data-driven modelling of streamlined weirs.

Keyphrases: Deep Convolutional Neural Network, Discharge prediction, Gated Recurrent Unit, Streamlined weirs, deep learning, machine learning

@booklet{EasyChair:7761,
  author    = {Weibin Chen and Danial Sharifrazi and Guoxi Liang and Shahab S. Band and Kwok Wing Chau and Amir Mosavi},
  title     = {Accurate Discharge Coefficient Prediction of Streamlined Weirs by Coupling Linear Regression and Deep Convolutional Gated Recurrent Unit},
  howpublished = {EasyChair Preprint 7761},
  year      = {EasyChair, 2022}}