eduzhai > Applied Sciences > Engineering >

Modeling and Uncertainty Analysis of Groundwater Level Using Six Evolutionary Optimization Algorithms Hybridized with ANFIS SVM and ANN

  • Save

... pages left unread,continue reading

Document pages: 42 pages

Abstract: In the present study, six meta-heuristic schemes are hybridized withartificial neural network (ANN), adaptive neuro-fuzzy interface system (ANFIS),and support vector machine (SVM), to predict monthly groundwater level (GWL),evaluate uncertainty analysis of predictions and spatial variation analysis.The six schemes, including grasshopper optimization algorithm (GOA), cat swarmoptimization (CSO), weed algorithm (WA), genetic algorithm (GA), krillalgorithm (KA), and particle swarm optimization (PSO), were used to hybridizefor improving the performance of ANN, SVM, and ANFIS models. Groundwater level(GWL) data of Ardebil plain (Iran) for a period of 144 months were selected toevaluate the hybrid models. The pre-processing technique of principal componentanalysis (PCA) was applied to reduce input combinations from monthly timeseries up to 12-month prediction intervals. The results showed that theANFIS-GOA was superior to the other hybrid models for predicting GWL in thefirst piezometer and third piezometer in the testing stage. The performance ofhybrid models with optimization algorithms was far better than that ofclassical ANN, ANFIS, and SVM models without hybridization. The percent ofimprovements in the ANFIS-GOA versus standalone ANFIS in piezometer 10 were14.4 , 3 , 17.8 , and 181 for RMSE, MAE, NSE, and PBIAS in the training stageand 40.7 , 55 , 25 , and 132 in testing stage, respectively. The improvementsfor piezometer 6 in train step were 15 , 4 , 13 , and 208 and in the test stepwere 33 , 44.6 , 16.3 , and 173 , respectively, that clearly confirm thesuperiority of developed hybridization schemes in GWL modeling. Uncertaintyanalysis showed that ANFIS-GOA and SVM had, respectively, the best and worstperformances among other models. In general, GOA enhanced the accuracy of theANFIS, ANN, and SVM models.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×