A split-step particle swarm optimization algorithm in river stage forecasting

Chau, Kwok Wing

doi:10.1016/j.jhydrol.2007.09.004

Cited by 113 publications

(51 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Chau [104] proposed a splitstep PSO algorithm which combines particle swarm optimization for global search and LevenbergMarquardt algorithm for fast local search in river stage forecasting. The algorithm provided more improved results in terms of computation time and accuracy.…”

Section: Performance Assessmentmentioning

confidence: 99%

Machine Learning Models Coupled with Variational Mode Decomposition: A New Approach for Modeling Daily Rainfall-Runoff

2018

View full text Add to dashboard Cite

Accurate modeling for nonlinear and nonstationary rainfall-runoff processes is essential for performing hydrologic practices effectively. This paper proposes two hybrid machine learning models (MLMs) coupled with variational mode decomposition (VMD) to enhance the accuracy for daily rainfall-runoff modeling. These hybrid MLMs consist of VMD-based extreme learning machine (VMD-ELM) and VMD-based least squares support vector regression (VMD-LSSVR). The VMD is employed to decompose original input and target time series into sub-time series called intrinsic mode functions (IMFs). The ELM and LSSVR models are selected for developing daily rainfall-runoff models utilizing the IMFs as inputs. The performances of VMD-ELM and VMD-LSSVR models are evaluated utilizing efficiency and effectiveness indices. Their performances are also compared with those of VMD-based artificial neural network (VMD-ANN), discrete wavelet transform (DWT)-based MLMs (DWT-ELM, DWT-LSSVR, and DWT-ANN) and single MLMs (ELM, LSSVR, and ANN). As a result, the VMD-based MLMs provide better accuracy compared with the single MLMs and yield slightly better performance than the DWT-based MLMs. Among all models, the VMD-ELM and VMD-LSSVR models achieve the best performance in daily rainfall-runoff modeling with respect to efficiency and effectiveness. Therefore, the VMD-ELM and VMD-LSSVR models can be an alternative tool for reliable and accurate daily rainfall-runoff modeling.

show abstract

Section: Performance Assessmentmentioning

confidence: 99%

Machine Learning Models Coupled with Variational Mode Decomposition: A New Approach for Modeling Daily Rainfall-Runoff

2018

View full text Add to dashboard Cite

show abstract

“…Recent research has shown that the PSO approach has many computational advantages over traditional evolutionary computing (Chau 2007). However, the drawback of premature convergence degrades its performance and reduces its global search ability.…”

Section: Methods Of Model Calibration and Validationmentioning

confidence: 99%

Hydrological Impacts of Climate Change Simulated by HIMS Models in the Luanhe River Basin, North China

Jiang

Liu

2014

Water Resour Manage

View full text Add to dashboard Cite

This paper applied a HIMS (hydroinformatic modeling system) model to simulate streamflow in the Luanhe River Basin. This model was compared with SIMHYD and XAJ models for eight sub-basins of the Luanhe River. The results showed HIMS model performed better than SIMHYD and XAJ models, in these areas. We then investigated the streamflow response to climate changes in the different sub-basins. Twenty hypothetical climate change scenarios (perturbed temperatures and precipitation) were used to test the sensitivity of HIMS model simulated annual and mean monthly streamflow. Our results demonstrated that: (i) the annual streamflow was positively related to precipitation, and there was a negative relationship between streamflow and temperature for all the eight sub-basins; (ii) in all sub-basins, the relationship of annual streamflow change to precipitation change was highly non-linear, but the relationship of annual streamflow change with temperature change was approximately linear; (iii) the annual streamflow response to precipitation change was more sensitive when increasing than decreasing; (iv) the annual streamflow response to climate change was more sensitive in the Xingzhouhe River sub-basin, followed by the Wuliehe River sub-basin, and the Sahe River sub-basin was least sensitive; (iv) there were few differences in inner-streamflow response to climate change in the Laoniuhe, Yimatuhe, and Yixunhe Rivers. But for other rivers, when the temperature changed, larger streamflow differences happened in winter and summer; when the precipitation decreased or was unchanged, the larger differences happened in winter months, and when the precipitation increased, larger differences happened in winter and summer.

show abstract

“…Since Gill et al (2006) used PSO for parameter estimation in hydrology, many researchers have presented various types of PSO for hydrologic model calibration (Jiang et al, 2007(Jiang et al, , 2010Zhang et al, 2009;Kuok et al, 2010;Kraue et al, 2011). PSO approach has many computational advantages over traditional evolutionary computing, such as rapid convergence (Jiang et al, , 2007Chau, 2007). Nonetheless, similar to general heuristic method, PSO also has the drawback of premature convergence, which degrades its performance and global search ability.…”

Section: Introductionmentioning

confidence: 99%

Rainfall-runoff modeling, parameter estimation and sensitivity analysis in a semiarid catchment

Jiang

Liu

et al. 2015

Environmental Modelling & Software

View full text Add to dashboard Cite

a b s t r a c tWe present a study on the Hydro-Informatic Modelling System (HIMS) rainfall-runoff model for a semiarid region. The model includes nine parameters in need of calibration. A master-slave swarms, shuffling evolution algorithm based on self-adaptive dynamic particle swarm optimization (MSSE-SDPSO) is proposed to derive model parameters. In comparison with SCE-UA, PSO, MSSE-PSO and MSSE-SPSO algorithms, MSSE-SDPSO has faster convergence and more stable performance. The model is used to simulate discharge in the Luanhe River basin, a semiarid region. Compared with the SimHyd and SMAR models, HIMS model has the highest Nash-Sutcliffe efficiencies (NSE) and smallest relative errors (RE) of volumetric fitness for the periods of calibration and verification. In addition, the studies indicate that the HIMS model with all-gauge data improves runoff prediction compared with single-gauge data. A distributed HIMS model performs better than a lumped one. Finally, the Morris method is used to analyze model parameters sensitivity for the objective functions NSE and RE.

show abstract

A split-step particle swarm optimization algorithm in river stage forecasting

Cited by 113 publications

References 20 publications

Machine Learning Models Coupled with Variational Mode Decomposition: A New Approach for Modeling Daily Rainfall-Runoff

Machine Learning Models Coupled with Variational Mode Decomposition: A New Approach for Modeling Daily Rainfall-Runoff

Hydrological Impacts of Climate Change Simulated by HIMS Models in the Luanhe River Basin, North China

Rainfall-runoff modeling, parameter estimation and sensitivity analysis in a semiarid catchment

Contact Info

Product

Resources

About