基于改进粒子群算法的四辊轧机机座结构优化设计  

Optimal Structural Design of 4-Roller Rolling Mill Frame Based on an IPSO Algorithm

作　　者：魏锋涛[1] 宋俐[1] 李言[1] 刘鹏[1] 

机构地区：[1]西安理工大学,西安710048

出　　处：《中国机械工程》2012年第19期2361-2365,共5页China Mechanical Engineering

基　　金：国家自然科学基金资助项目(60903124);陕西省重点学科建设专项资金资助项目(102-00X903)

摘　　要：为了获得四辊轧机的合理结构参数与最佳工作性能,在满足结构、强度、刚度等约束条件下,以机座弹跳值最小为目标函数,建立四辊轧机机座结构优化设计数学模型。针对常规优化方法和标准粒子群算法求解该类复杂机械系统结构优化设计问题存在的诸多缺陷,提出采用Logistic映射对种群进行混沌初始化、非线性地动态改进惯性权值和非线性地动态调整加速因子三种策略对标准粒子群算法进行改进,结合改进的惩罚函数法对四辊轧机机座结构进行优化设计,求解结果验证了所提方法的正确性和有效性。In order to get some reasonable parameters and best performance of a structure,taking the minimum bounce value of frame as an objective,a mathematical model for optimal design of the 4 --roller rolling mill frame was established under the conditions that the frame met the requirements for structure,strength, stiffness and others. Considering the defects existed in the conventional optimal algorithm and standard particle swarm optimization algorithm used for dealing with complex me- chanical systems,an IPSO algorithm was proposed which included three kinds of strategies, Logistic map initiating the population chaotically, nonlinear dynamic improving inertial weights and adjusting accelerated factor. The structural parameters of a 4--roller rolling mill frame was designed optimally by using an improved penalty function method and an IPSO algorithm. The results verify that the improved algorithm proposed is correct and effective.

关 键 词：四辊轧机机座 结构参数 优化设计 改进粒子群算法 惩罚函数法 

分 类 号：TH122[机械工程—机械设计及理论]