Size and shape optimization of trusses under dead and moving loads using genetic algorithm

Size and Shape Optimization of Trusses under Dead and Moving Loads Using Genetic Algorithm Various artificial intelligence tools are used to solve engineering problems for last three or four decade. Genetic algorithm (GA) is one of the methods to take most attentions among these tools. The GA is stochastic global search method that imitates the process of natural biological evolution. Based on the concept of genetics, GA simulates the evolutionary process numerically analogous to genes in genetics, GA represent the parameters encoded in a string for a given problem. In contrast to mathematical optimization methods, which find the optimal solution with a sequential search, a set of points each representing the solution of problem or system in design space is used to search for the optimal solution in GA. GA tries to find the best solution ensuring desired criteria among these points. In this study, the optimum size and shape design of 3D trusses under multiple loads and the minimum weight design of bridges trusses under moving load is performed using GA with value encoding, mixed crossover or adaptive mutation and crossover operators. A program is coded for the purpose. Stress, stability, and maximum displacements are considered as the design criteria obeying TS 648 / AISC. Numerical examples from the technical literature are solved to show the performance of the developed algorithm and results compared. Later, the size and shape optimization several 3D truss systems are performed. And finally, three bridge truss systems are optimized under moving load. Conclusions drawn from the study and the suggestions for the future research are summarized. It can be concluded that to use size and shape optimization together to get the minimum weight of the 3D trusses is more appropriate than to use size optimization only. Adaptive mutation and crossover operators increase performance of GA, and GA performs well to find minimum weight of bridges trusses under moving load. Keywords: Size and Shape Optimization, Genetic Algorithm, Value Encoding, Adaptive Operators, Bridge Trusses, Moving Load. VI