Rainfall-runoff modelling using improved machine learning methods: Harris hawks optimizer vs. particle swarm optimization

Tikhamarine, Yazid; Souag-Gamane, Doudja; Ahmed, Ali Najah; Sammen, Saad Sh.; Kisi, Özgür; Huang, Yuk Feng; El-Shafie, Ahmed

doi:10.1016/j.jhydrol.2020.125133

Cited by 108 publications

(40 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several kinds of machine learning model that can be used to attain a precise estimation on the short-term runoff have been shown in several pieces of research. For example, the support vector machine [16][17][18] and the random forest regressor [19,20]. As a subset of machine learning, deep learning, which is mainly represented by the artificial neural network (ANN) techniques, is of great interest nowadays due to the booming computer science and algorithms [21].…”

Section: Introductionmentioning

confidence: 99%

Impact of Input Filtering and Architecture Selection Strategies on GRU Runoff Forecasting: A Case Study in the Wei River Basin, Shaanxi, China

Wang

Liu

Yue

et al. 2020

Water

View full text Add to dashboard Cite

A gated recurrent unit (GRU) network, which is a kind of artificial neural network (ANN), has been increasingly applied to runoff forecasting. However, knowledge about the impact of different input data filtering strategies and the implications of different architectures on the GRU runoff forecasting model’s performance is still insufficient. This study has selected the daily rainfall and runoff data from 2007 to 2014 in the Wei River basin in Shaanxi, China, and assessed six different scenarios to explore the patterns of that impact. In the scenarios, four manually-selected rainfall or runoff data combinations and principal component analysis (PCA) denoised input have been considered along with single directional and bi-directional GRU network architectures. The performance has been evaluated from the aspect of robustness to 48 various hypermeter combinations, also, optimized accuracy in one-day-ahead (T + 1) and two-day-ahead (T + 2) forecasting for the overall forecasting process and the flood peak forecasts. The results suggest that the rainfall data can enhance the robustness of the model, especially in T + 2 forecasting. Additionally, it slightly introduces noise and affects the optimized prediction accuracy in T + 1 forecasting, but significantly improves the accuracy in T + 2 forecasting. Though with relevance (R = 0.409~0.763, Grey correlation grade >0.99), the runoff data at the adjacent tributary has an adverse effect on the robustness, but can enhance the accuracy of the flood peak forecasts with a short lead time. The models with PCA denoised input has an equivalent, even better performance on the robustness and accuracy compared with the models with the well manually filtered data; though slightly reduces the time-step robustness, the bi-directional architecture can enhance the prediction accuracy. All the scenarios provide acceptable forecasting results (NSE of 0.927~0.951 for T + 1 forecasting and 0.745~0.836 for T + 2 forecasting) when the hyperparameters have already been optimized. Based on the results, recommendations have been provided for the construction of the GRU runoff forecasting model.

show abstract

Section: Introductionmentioning

confidence: 99%

Impact of Input Filtering and Architecture Selection Strategies on GRU Runoff Forecasting: A Case Study in the Wei River Basin, Shaanxi, China

Wang

Liu

Yue

et al. 2020

Water

View full text Add to dashboard Cite

show abstract

“…In addition to that, recently, different advanced optimization techniques are proposed as robust techniques to be used in searching for the best solution in dealing with water resources issues [26]. For example, the waterdrop optimization technique [38], the whale optimization algorithm [39], the ant lion optimization algorithm [40], the nomadic people optimization algorithm [41], the Harris hawks optimization algorithm [42], and the grey wolf optimization technique [27]. Therefore, future work could be carried out to develop these recent optimization techniques and explore their performances in optimizing the release from the reservoir to meet the downstream demand.…”

Section: Reliability and Risk Analysismentioning

confidence: 99%

Optimizing the Operation Release Policy Using Charged System Search Algorithm: A Case Study of Klang Gates Dam, Malaysia

et al. 2021

Self Cite

View full text Add to dashboard Cite

In planning and managing water resources, the implementation of optimization techniques in the operation of reservoirs has become an important focus. An optimal reservoir operating policy should take into consideration the uncertainty associated with uncontrolled reservoir inflows. The charged system search (CSS) algorithm model is developed in the present study to achieve optimum operating policy for the current reservoir. The aim of the model is to minimize the cost of system performance, which is the sum of square deviations from the distinction between the release of the target and the actual demand. The decision variable is the release of a reservoir with an initial volume of storage, reservoir inflow, and final volume of storage for a given period. Historical rainfall data is used to approximate the inflow volume. The charged system search (CSS) is developed by utilizing a spreadsheet model to simulate and perform optimization. The model gives the steady-state probabilities of reservoir storage as output. The model is applied to the reservoir of Klang Gates for the development of an optimal reservoir operating policy. The steady-state optimal operating system is used in this model.

show abstract

“…Soft computing techniques have successfully used in the last three decades to solve different complex hydrological problems (Daliakopoulos et al, 2005;Ehteram et al, 2021;Hadi et al, 2019;Kaloop et al, 2017;Malik et al, 2020b;Parsaie et al, 2015;Sammen et al, 2017;Singh et al, 2018;Tikhamarine et al, 2020c;Yaseen et al, 2020b;Young et al, 2015). For water level prediction, several techniques have been used such as Artificial Neural Networks (ANN) (Alvisi et al, 2006), Autoregressive Integrated Moving Average (ARIMA) (Reza et al, 2018;Sihag et al, 2020;Xu et al, 2019), and Support Vector Machine (SVM) (Khan & Coulibaly, 2006;Liong & Sivapragasam, 2002).…”

Section: Introductionmentioning

confidence: 99%

“…Evaluation of soft computing performance conducted by Firat (2008), Toriman et al (2009), and Khairuddin et al (2019) acknowledged superiority over statistical and time series methods for flood forecasting. Besides, the soft computing/ machine learning (ML) models received several practical applications in diverse fields like the prediction of solar radiation (Qin et al, 2018;Wang et al, 2016Wang et al, , 2017b, evaporation modeling (Adnan et al, 2019;Ashrafzadeh et al, 2020;Malik et al, 2017Malik et al, , 2018Wang et al, 2017aWang et al, , 2017c, rainfall-runoff forecasting (Malik et al, 2020b;Singh et al, 2018;Tikhamarine et al, 2020c), reference evapotranspiration estimation (Malik et al, 2019a;Mohamadi et al, 2020;Tikhamarine et al, 2019Tikhamarine et al, , 2020aTikhamarine et al, , 2020b, meteorological and hydrological drought prediction (Malik et al, 2019c(Malik et al, , 2020a(Malik et al, , 2021a(Malik et al, , 2021b(Malik et al, , 2021cMalik & Kumar, 2020), and simulation of seepage flow through embankment dam (Rehamnia et al, 2021).…”

Section: Introductionmentioning

confidence: 99%

Prediction of daily water level using new hybridized GS-GMDH and ANFIS-FCM models

Ebtehaj

Sammen

Sidek

et al. 2021

Engineering Applications of Computational Fluid Mechanics

Self Cite

View full text Add to dashboard Cite

Accurate prediction of water level (WL) is essential for the optimal management of different water resource projects. The development of a reliable model for WL prediction remains a challenging task in water resources management. In this study, novel hybrid models, namely, Generalized Structure-Group Method of Data Handling (GS-GMDH) and Adaptive Neuro-Fuzzy Inference System with Fuzzy C-Means (ANFIS-FCM) were proposed to predict the daily WL at Telom and Bertam stations located in Cameron Highlands of Malaysia. Different percentage ratio for data division i.e. 50%-50% (scenario-1), 60%-40% (scenario-2), and 70%-30% (scenario-3) were adopted for training and testing of these models. To show the efficiency of the proposed hybrid models, their results were compared with the standalone models that include the Gene Expression Programming (GEP) and Group Method of Data Handling (GMDH). The results of the investigation revealed that the hybrid GS-GMDH and ANFIS-FCM models outperformed the standalone GEP and GMDH models for the prediction of daily WL at both study sites. In addition, the results indicate the best performance for WL prediction was obtained in scenario-3 (70%-30%). In summary, the results highlight the better suitability and supremacy of the proposed hybrid GS-GMDH and ANFIS-FCM models in daily WL prediction, and can, serve as robust and reliable predictive tools for the study region.

show abstract

Rainfall-runoff modelling using improved machine learning methods: Harris hawks optimizer vs. particle swarm optimization

Cited by 108 publications

References 40 publications

Impact of Input Filtering and Architecture Selection Strategies on GRU Runoff Forecasting: A Case Study in the Wei River Basin, Shaanxi, China

Impact of Input Filtering and Architecture Selection Strategies on GRU Runoff Forecasting: A Case Study in the Wei River Basin, Shaanxi, China

Optimizing the Operation Release Policy Using Charged System Search Algorithm: A Case Study of Klang Gates Dam, Malaysia

Prediction of daily water level using new hybridized GS-GMDH and ANFIS-FCM models

Contact Info

Product

Resources

About