Performance evaluation of ANN-WOA and ANN-BWO simulation-optimization methods in predicting daily runoff (case study: Jelogir station in Karkheh watershed)

Document Type : Research Paper


1 Assistant Professor, Department of Civil Engineering, Science and Research Branch, Islamic Azad University, Tehran, Iran

2 Department of Civil Engineering, Science and Research Branch, Islamic Azad University,Tehran, Iran

3 Department of Range and Watershed Management, Urmia University, Urmia, Iran

4 Department of Water Engineering, Center of Excellence in Hydroinformatics, Faculty of Civil Engineering, University of Tabriz, Tabriz, Iran



Rainfall-runoff modeling is one of the methods of estimating runoff and a suitable tool for studying hydrological processes, evaluating water resources and watershed management. But the complexity and non-linear nature of the rainfall-runoff process and the unknown effect of the factors on each other and finally on the outflow of the basin make modeling more difficult.

Results and Discusspn

In this research, the correlation coefficient was used to find the relationship between the input and output variables. The results of the research show that the flow rate with a one-day delay had the highest correlation with the output flow rate. Also, after choosing the best combination of the input model from the artificial neural network to predict the process Precipitation-runoff was used and black spider and whale algorithms were used to optimize the weighting and bias coefficients. The final results showed that all the models had a very good performance in forecasting and were able to predict the result of the model in a single mode. Finally, a box diagram and time series and data dispersion were used.


According to the RMSE criteria, it can be said that the ANN-WOA model has the best performance in predicting the rainfall-runoff process. Also, all the mentioned models showed a very good performance in the forecasting process. Based on this, the ANN-WOA model has been able to improve the accuracy of a single model by 32.4%, the ANN-BWO model by 27.6% and the WANN network by 22.4%.


Main Subjects

  • Receive Date: 02 October 2022
  • Revise Date: 16 August 2023
  • Accept Date: 17 September 2023
  • Publish Date: 17 September 2023