Minimum Cost Design of Hydraulic Water Retaining Structure by using Coupled Simulation Optimization Approach

摘要

A linked Simulation-Optimization (S-O) model was developed to find the optimum design of Hydraulic Water Retaining Structure (HWRS) constructed on permeable soils. The nonlinear relationship between seepage design variables can accurately and solely simulated by numerical methods. To increase the computational efficiency of the S-O model, the seepage numerical model was replaced by approximation simulator based on the Support Vector Machine (SVM) surrogate models. The surrogate models incorporated and highlighted the effects of hydraulic conductivity (k) and anisotropy ratio (k(y)/k(x)) on the optimum design of HWRS. The results revealed that reducing k and (k(y)/k(x)) values augments the optimum cost. The most effective seepage controller variables were the depth and inclination angle of the last cut-off. Increasing these variables effectively reduced the exit gradient value to the allowable limits. Also, the first and second last aprons (b(9), b(10)) were important to provide a sufficient cross section for HWRS to increase the stability of the HWRS against the overturning, flotation, sliding, etc. The evaluation of S-O technique demonstrated a good agreement between the predicted and simulated seepage characteristics of the optimum HWRS design. Hence, the S-O methodology is applicable to obtain an optimal and minimum cost HWRS design.