Estimating the solute transport parameters of the spatial fractional advection-dispersion equation using Bees Algorithm

• Published in: Journal of contaminant hydrology Vol. 203; pp. 51 - 61
• Main Author: Mehdinejadiani, Behrouz
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01.08.2017
• Subjects: Algorithms; equations; FracFit toolbox; Global optimum solution; Groundwater; Hydrology - methods; Inverse method; Meta-heuristic algorithms; Models, Theoretical; soil; Soil - chemistry; solutes; FracFit toolbox; Global optimum solution; Meta-heuristic algorithms; Inverse method
• ISSN: 0169-7722; 1873-6009; 1873-6009
• DOI: 10.1016/j.jconhyd.2017.06.004

This study represents the first attempt to estimate the solute transport parameters of the spatial fractional advection-dispersion equation using Bees Algorithm. The numerical studies as well as the experimental studies were performed to certify the integrity of Bees Algorithm. The experimental ones were conducted in a sandbox for homogeneous and heterogeneous soils. A detailed comparative study was carried out between the results obtained from Bees Algorithm and those from Genetic Algorithm and LSQNONLIN routines in FracFit toolbox. The results indicated that, in general, the Bees Algorithm much more accurately appraised the sFADE parameters in comparison with Genetic Algorithm and LSQNONLIN, especially in the heterogeneous soil and for α values near to 1 in the numerical study. Also, the results obtained from Bees Algorithm were more reliable than those from Genetic Algorithm. The Bees Algorithm showed the relative similar performances for all cases, while the Genetic Algorithm and the LSQNONLIN yielded different performances for various cases. The performance of LSQNONLIN strongly depends on the initial guess values so that, compared to the Genetic Algorithm, it can more accurately estimate the sFADE parameters by taking into consideration the suitable initial guess values. To sum up, the Bees Algorithm was found to be very simple, robust and accurate approach to estimate the transport parameters of the spatial fractional advection-dispersion equation. •The transport parameters of sFADE are estimated using the Bees Algorithm (BA).•The sFADE parameters resulted from BA are compared with those from FracFit.•The BA has very high practical reliabilities at all cases.•The results of BA are significantly more accurate than those of FracFit.•The BA showed the relative similar performances for all cases.