Assessing soft calculation methods in river suspended sediment estimation (Hassan Abad station of Tirah river)

Introduction

Rivers are always faced with the phenomenon of erosion and sediment transfer. Sediment transport in rivers is one of the most complex topics in river engineering and is always the focus of experts and water engineers. Various factors affect this phenomenon, which makes their analysis difficult. Statistical and regression models are the most common methods of analysis, which often provide erroneous results due to the linear solution of these phenomena. Therefore, they cannot model the sedimentation phenomenon with acceptable accuracy. Hydraulic models cannot always be trusted due to the need for a lot of data and sometimes the unavailability of the required data and the inaccuracy of the data due to human error for simulating sediments. Nowadays, fuzzy and neural intelligent conductor systems, due to their ability to solve complex and nonlinear phenomena, have found many applications in various water engineering problems, including sedimentation. The purpose of this research is to evaluate and compare adaptive neural fuzzy models (ANFIS), support vector machine (SVM) and GEP gene expression programming, GMDH data control group method in estimating the sediment load of Hasan Abad Station,

Methods

In this research, first, the long-term daily statistics of temperature, rainfall, average flow rate and sediment concentration of Hasan Abad hydrometric and sediment measuring station located on the main branch of the Tirah River were collected. Then, the data sufficiency test for analysis, checking the correlation between parameters of river discharge, precipitation, temperature with sediment discharge, and determining the long-term average of suspended sediment in the studied stations were performed. The design of the input parameter pattern can be based on the relationship between flow and sediment flow parameters, rainfall, temperature, flow, and sediment flow. Of course, considering that the mentioned parameters have a historical course, therefore, the design of the input patterns of soft computing models should be done based on time delays (like what is discussed in the analysis and forecasting of time series). Determining the most appropriate time delay of the input parameters in the modeling of discharge, sediment, temperature, and rainfall, then the appropriate design of the structure of the used soft calculation models was done. the next step, the estimation of sediment discharge using a support vector machine (SVM), gene expression programming (GEP), and fuzzy-adaptive neural system (Anfis) group method of GMDH data control and comparison of three data mining methods.

Results and Discussion

In this research, based on the statistical indicators of optimal model selection, the best performance of the SVR model has been obtained for model number 1. In this model, the R2 explanatory coefficient and the RMSE error obtained from the model are 0.96 and 0.0047, respectively. The coefficient of explanation R2 and the RMSE error of the models in predicting suspended sediment values in the test stage are 0.95 and 0.014 respectively for the ANFIS model and 0.50 and 4.97 respectively for the GEP model. Came. The best performance of the ANFIS model has been obtained for model number 1. In this model, the R2 explanation coefficient and the RMSE error obtained from the model are 0.95 and 0.014 tons per day. The coefficient of explanation R2 and the RMSE error of the models in predicting suspended sediment values in the test stage are 0.96, 0.0047 for the SVR model, and 0.50, 4.97 for the GEP model, respectively. was achieved The best performance of the GEP model has been obtained for pattern number 9. In this model, the R2 explanation coefficient and the RMSE error obtained from the model are 0.99 and 0.010 tons per day, respectively. The coefficient of explanation R2 and the RMSE error of the models in predicting the amount of suspended sediment in the test stage are respectively equal to 0.70, 0.015 for the ANFIS model and 0.78, 0.0185 tons respectively for the SVR model in The day has been achieved. The comparison of the results of three ANFIS, GEP, and SVR models indicates the superiority of the GEP model in predicting the amount of suspended sediment according to the input model 9.

Conclusion

According to the obtained results, it can be seen that the performance of the GEP model was better compared to other models. SVR and ANFIS models are ranked second and third. First, input pattern 1, which was selected as the best pattern for (ANFIS) and (SVM) models, was introduced as the input of the GMDH model. In the training and test, the values of R2 statistical indices are 0.94 and 0.99 respectively, the RMSE error value is 0.0079 and 0.0038 respectively, the MSE value is 0.000062 and 0.000015 respectively, and the MAPE values are respectively 0.007 and 0.003 were obtained. In the next step, input pattern 9, which was selected as the best pattern for the GEP model, was introduced as GMDH input. In the training and test, the values of R2 statistical indices are equal to 0.95 and 0.98 respectively, the RMSE error value is equal to 0.0077 and 0.0045 respectively, and the MSE value is equal to 0.0006 and 0.00002 respectively, and MAPE values are The order of 363 and 502 was obtained. The results show the acceptable performance of the models compared to the sediment rating curve. Also, the results showed the superiority of the model (GEP) with the highest determination coefficient R2 with a value of 0.99 and the lowest root mean square error RMSE in terms of tons per day with a value of 0.010. In this regard, the efficiency of the model (SVM) was somewhat better than the model (ANFIS). In the next step, the best -selected pattern of the (ANFIS), (SVM), and (GEP) models was used as the input of the GMDH model. The results showed the acceptable performance of the GMDH model with the highest coefficient of R2 determination equal to 0.99 and 0.98 and the lowest root mean square error equal to 0.0038 and 0.0045 tons per day, respectively. The obtained results showed that all four investigated data mining methods provide far better results than the sediment rating curve.