Truss optimization on shape and sizing with frequency constraints based on genetic algorithm
Truss optimization on shape and sizing with frequency constraints based on genetic algorithm
复制标题
DOI：
10.1007/s00466-004-0623-8
复制
发表时间：
2005-04
影响因子：
4.1
通讯作者：
Lingyun Wei;Zhao Mei;Guangming Wu;M. Guang
中科院分区：
文献类型：
--
作者：
Lingyun Wei;Zhao Mei;Guangming Wu;M. Guang
基于频率约束的桁架形状和尺寸优化是高度非线性的动态优化问题。节点坐标和横截面积这两种不同类型的设计变量的耦合往往会导致离散，而多个频率约束往往会给动态灵敏度分析带来困难。因此，最优准则法和数学规划等动态灵敏度复杂且容易陷入局部最优的方法难以解决。为了简单有效地解决桁架形状和尺寸优化问题，提出了一种小生境混合遗传算法（NHGA）。NHGA的目标是在新的混合架构的基础上，在提高挖掘能力的同时防止过早收敛。采用小生境技术和自适应参数调整来保持种群多样性，防止过早收敛；采用单纯形搜索来增强遗传算法的局部搜索能力。该算法有效地缓解了遗传算法的过早收敛，改善了遗传算法的弱挖掘能力。通过几个典型桁架优化算例，验证了NHGA求解多频率约束下桁架形状和尺寸优化的有效性、有效性和可靠性。
Truss optimization on shape and sizing with frequency constraints are highly nonlinear dynamic optimization problems. Coupling of two different types of design variables, nodal coordinates and cross-sectional areas, often lead to divergence while multiple frequency constraints often cause difficult dynamic sensitivity analysis. So optimal criteria method and mathematical programming, which need complex dynamic sensitivity and are easily trapped into the local optima, are difficult to solve the problems. To solve the truss shape and sizing optimization simply and effectively, a Niche Hybrid Genetic Algorithm (NHGA) is proposed. The objective of NHGA is to enhance the exploitation capacities while preventing the premature convergence simultaneously based on the new hybrid architecture. Niche techniques and adaptive parameter adjustment are used to maintain population diversity for preventing the premature convergence while simplex search is used to enhance the local search capacities of GAs. The proposed algorithm effectively alleviates premature convergence and improves weak exploitation capacities of GAs. Several typical truss optimization examples are employed to demonstrate the validity, availability and reliability of NHGA for solving shape and sizing optimization of trusses with multiple frequency constraints.