Đang chuẩn bị nút TẢI XUỐNG, xin hãy chờ
Tải xuống
In this paper, the optimal sizing of truss structures is solved using a novel evolutionary-based optimization algorithm. The efficiency of the proposed method lies in the combination of global search and local search, in which the global move is applied for a set of random solutions whereas the local move is performed on the other solutions in the search population. | Vietnam Journal of Mechanics, VAST, Vol. 38, No. 4 (2016), pp. 307 – 317 DOI:10.15625/0866-7136/7476 AN EVOLUTIONARY-BASED OPTIMIZATION ALGORITHM FOR TRUSS SIZING DESIGN Pham Hoang Anh National University of Civil Engineering, Hanoi, Vietnam E-mail: anhpham.nuce@gmail.com Received November 28, 2015 Abstract. In this paper, the optimal sizing of truss structures is solved using a novel evolutionary-based optimization algorithm. The efficiency of the proposed method lies in the combination of global search and local search, in which the global move is applied for a set of random solutions whereas the local move is performed on the other solutions in the search population. Three truss sizing benchmark problems with discrete variables are used to examine the performance of the proposed algorithm. Objective functions of the optimization problems are minimum weights of the whole truss structures and constraints are stress in members and displacement at nodes. Here, the constraints and objective function are treated separately so that both function and constraint evaluations can be saved. The results show that the new algorithm can find optimal solution effectively and it is competitive with some recent metaheuristic algorithms in terms of number of structural analyses required. Keywords: Structural optimization, evolutionary-based optimization, metaheuristics, truss structure, sizing optimization. 1. INTRODUCTION Truss optimization is one of the most popular design problems and has been an extensive research area both in modeling and development of optimization methods. Often the weight of truss structure is to be minimized subject to stress, displacement, and/or natural frequency constraints. This optimization task is in general difficult to solve because of non-linear constraints and non-convex feasible region. This means that the convergence of traditional gradient-based optimization methods cannot be ensured [1]. Metaheuristics such as genetic algorithms, particle
