Layout Optimization of Double-Layer Grids Using Modified Genetic Algorithm Based on Fuzzy Inference System

Abstract

Inweight optimizitaion of double-layer grids, various parameters such as the members cross-sectional areas, the height between the two layers, the structure meshing in two directions and topology of the structure should be considered. In this study, for simultaneous optimization of size, shape and topology of double-layer grids, genetic algorithm is employed and is modified based on fuzzy inference system. First, to efficiently search in design space at each stage, some solutions are generated in the neighborhood of the best sample, which enhances searching operation in the neighborhood of the optimum solution. Then, in order to achieve the possible solutions, the penalties for violation of constraints and the number of violated constraints are considered to choose the next generation. The value of objective function and the values of genetic algorithm parameters have a great effect on the result of the algorithm. In order to adaptive setting of these parameters, the fuzzy inference system is employed. The efficiency of these improvements has been confiremd by presenting some examples of truss structures and comparison with the other algorithms.