Fuzzy logic applied to the modeling of the transfer of water in soils / LÃgica fuzzy aplicada à modelagem da transferÃncia de Ãgua em solos

AUTOR(ES)
DATA DE PUBLICAÇÃO

2009

RESUMO

Modeling water movement in the unsaturated soil requires a great number of parameters and variables such as the initial soil-water content, the saturated soil-water content, and the saturated hydraulic conductivity, which can relatively easily be assessed. Other functions such as hydraulic conductivity and matric potential which are nonlinear functions of the actual soil-water content are more difficult to estimate. Monodimensional water flow in the unsaturated zone is usually modeled by a nonlinear partial differential equation, known as Richardsâs equation. Since this equation can only be solved analytically for some special cases, a numerical algorithm is a way to find approximate solutions. The success of numerical models to describe the dynamics of water in soil is closely related to the precision with which the soil-water physical parameters are determined. This has been the major challenge in the use of numerical models, because in general, these parameters are difficult to determine and present high spatial variability in soil. So, make it necessary to develop and use methodologies that incorporate, in an appropriate way, the uncertainties inherent to the movement of water in soils. In this study, fuzzy models are used as an alternative to represent the water flow in the unsaturated zone of the soil. Two fuzzy logic-based models were proposed to simulate the process of water redistribution in the soil. The principle of these models is a system based on fuzzy rules of Mamdani type. The training set was obtained by numerical solution of the Richardsâs equation by the method of finite differences (MDF) and was used to formulate two fuzzy rule-based models. The rules are based on the moisture content of the adjacent soil layers. Among the advantages of the fuzzy rule-based model developed in this work are its simplicity and low computational cost. The performance results modeled by fuzzy system is assessed by comparing the evolution of the soil-water volumetric profiles over time and with those obtained through numerical simulation of the Richardsâs equation under two different conditions of bottom boundary and to three soils with different hydrodynamic characteristics. The results obtained by the use of fuzzy rule-based models showed a satisfactory reproduction of values when compared with the numerical solution of the Richardsâs equation. Each of these models was adjusted appropriately for each case studied in this study. This fact confirmed the possibility of the fuzzy logic is perfect possible to model other cases, adopting the same methodology proposed and applied in this study.

ASSUNTO(S)

mamdani models fuzzy logic engenharia nuclear fuzzy rule-based models modelo de mamdani lÃgica fuzzy fluxo nÃo saturado energia nuclear mÃtodo das diferenÃas finitas (mdf) unsaturated flow

ACESSO AO ARTIGO

Documentos Relacionados