水电设备加工用机床立柱优化设计

Optimization design of machine tool column for hydropower equipment processing

针对现有立柱结构优化设计中模型复杂、计算量大及求解时间长等问题,提出一种结合拓扑优化、灵敏度分析、径向基函数(RBF)神经网络模型和带精英策略的非支配排序遗传算法(NSGA-Ⅱ)的组合优化设计方法。采用拓扑优化方法明确加工中心立柱材料的最佳分布,并进行模型修复;以立柱开孔、开槽和横竖筋板尺寸为输入参数,立柱质量、最大变形、最大等效应力和1阶固有频率作为目标进行灵敏度分析,筛选出6个关键尺寸作为优化设计变量。采用拉丁超立方抽样和有限元分析获得试验样本和结果,构建RBF神经网络优化模型并采用遗传算法求解。分析结果表明:优化后立柱质量减轻8.8%,1阶固有频率提升5%,立柱最大应力减小36%,立柱静动态特性都有所提高。

To solve the issues of complex models,vast computing,and long processing time in the existing column structure optimization design,a combined optimization design method integrating topological optimization,sensitivity analysis,Radial Basis Function(RBF)neural network models,and Non-dominated Sorting Genetic Algorithm II with elitist strategy(NSGA-II)is proposed.Using topology optimization to determine the optimal distribution of materials for machining center columns and conducting model repairs;Sensitivity analysis was conducted using column openings,slots,and dimensions of horizontal and vertical reinforcement plates as input parameters,with column mass,maximum deformation,maximum equivalent stress,and first-order natural frequency as objectives.Six key dimensions were selected as optimization design variables.Latin hypercube sampling and finite element analysis were used to obtain experimental samples and results,and an RBF neural network optimization model was constructed and solved using genetic algorithm.The analysis results show that after optimization,the weight of the column is reduced by 8.8%,the first-order natural frequency is increased by 5%,the maximum stress of the column is reduced by 36%,and the static and dynamic characteristics of the column are improved.