基于改进粒子群的薄壁变截面刚架临界载荷优化算法

Optimization algorithm of critical load of a thin-walled frame with variable cross-section based on improved particle swarm

针对大型变截面薄壁结构的稳定问题,以一类任意约束对称结构受非对称载荷的单跨刚架为研究对象,结构拆分为相关铁木辛柯(Timoshenko)梁,结合差分原理和最优化方法,以每段刚架的每个离散点挠度、临界载荷、轴力、剪力和梁端弯矩为设计变量,建立求解满足边界条件的非线性差分方程模型,提出基于优胜劣汰粒子更新的粒子群(IPSO)临界载荷优化算法。运用JAVA编程语言编制对应优化程序,分析典型算例并核实ABAQUS仿真结果。研究表明,本文提出的优化算法获得了有效的变形位型和高精度的临界载荷计算,能更好地描述刚架受力下位型和载荷的力学关系,进一步为工程设计与分析提供支持。

In order to study the stability of a large thin-walled structure, a kind of single-bay frame in symmetrical structure with arbitrary constraints under asymmetric loadings has been built.The frame is modelled as a number of Timoshenko beams.Based on the finite difference method and optimization method, using nodal deflections, critical load,axial force,shear force and bending moment as the design variables,nonlinear differential equations are built.An optimization algorithm of critical load based on improved particle swarm with competitive thought is put forward.JAVA computer language is used to develop the optimization procedure and ABAQUS used to obtain the simulation results of the example.Fewer design variables are needed to get effective position and critical load with high precision.A mechanical relation of position and loading on the frame could be better described to give a further support for engineering design and analysis.