Application of a Genetic Algorithm for the Optimization of a Location and Inclination Angle of a Cut-Off Wall for Anisotropic Foundations Under Hydraulic Structures

Abstract

A genetic algorithm technique, integrated with a numerical model (finite element method), was used in this study to compute the optimal cut-off location and angle of inclination for barrages constructed on homogenous anisotropic soil foundations. The exit gradient function is minimized as an objective function in the problem formulation. Constraints included uplift pressure functions and safety factors taking account of a minimum concrete floor thickness. Various degrees of anisotropy and different values of a range of heads have been studied. Around 1400 different situations (cases) were simulated using SEEP2D GMS software. Statistical nonlinear regression models were used to predict the pressure head and exit gradient for anisotropic soil foundations under various degrees of inclination (θ), relative location (b₁/b) and relative depth (d/b). A genetic algorithm optimization model was also performed using Matlab. The study results indicated that the computed optimum cut-off locations and relative inclination angles were affected by changing the anisotropic ratio and relative cut-off depth. For isotropic soil foundations, for cases with a cut-off ratio (d/b) of up to 0.4, the optimum inclination angle ranged between 59° and 68° for upstream locations. In other relative cut-off depth cases, the cut-off can work efficiently when situated in the last third of the concrete floor foundation, this considered the optimum location. For anisotropic soil, optimum locations tended to be upstream for most d/b values.