基于非概率可靠性的离散变量结构优化设计  

Optimization Design of Structures with Discrete Variables Based on Non-probabilistic Reliability

作　　者：张思才[1] 张方晓[1] ZHANG Si-cai ZHANG Fang-xiao(Institute of Systems Engineering, CAEP, Mianyang 621999, China)

机构地区：[1]中国工程物理研究院总体工程研究所,四川绵阳621999

出　　处：《装备环境工程》2017年第7期61-64,共4页Equipment Environmental Engineering

摘　　要：目的针对工程结构设计中因材料供应规范限制使得设计变量不能连续取值以及材料性能波动、外界载荷偏差等不确定性问题,探索研究非概率可靠性在离散变量结构优化设计中的应用。方法以桁架结构杆件截面直径为设计变量、结构质量极小化为目标,考虑材料强度极限与外部载荷的不确定性,建立具有非概率可靠性指标约束的离散变量结构可靠性优化模型,并采用遗传算法求解离散变量结构优化设计问题。结果求解获得了三杆、十杆超静定桁架杆件截面直径最优组合,得到的优化结果相较确定性优化结果更具鲁棒性。结论非概率可靠性模型对于解决工程中设计变量具有离散性和不确定性的优化设计问题具有可行性。Objective To explore application of non-probability reliability in optimization design of structure with the discrete variables to solve issues such as failure of continues valuing in design variables due to limit of material supply specification, fluctuation of material performance, deviation of external load, etc. Methods The nondeterminacy of material＇s ultimate strength and external load was considered with cross-sectional diameter of trusswork rod pieces as design variable, with minimizing structural weight as target to establish a reliable optimization model for discrete variable structure of non-probability reliability index constrain. The genetic algorithm was used to solve the problems of truss optimization design with the discrete variables. Results Optimal combination of cross-sectional diameters of three-rod and ten-rod indeterminate truss rod pieces was obtained. The optimization result obtained was more robust than that of the deterministic optimization. Conclusion Non probabilistic reliability model is feasible for solving discreteness and non-probability in optimization design of structures with discrete variables and the uncertainty.

关 键 词：非概率 离散变量 结构优化 可靠性 

分 类 号：TJ02[兵器科学与技术—兵器发射理论与技术]