Optimization Design of Hoist Towers Based on Dynamic Kriging Model with Double Point Adding Criterion

Abstract

Abstract: In order to minimize the weight of hoist towers under the conditions of satisfying the strength requirements and solve the problems that traditional Kriging model could hardly guarantee the global accuracy and local accuracy simultaneously, a dynamic Kriging model with double point adding criterion was put forward. Based on the model and artificial bee colony algorithm, the hoist towers were optimized. The sensitive parameters determined by global sensitivity analyses of hoist towers were regarded as design variables, then the initial Kriging model was established by sample data obtained by the Latin hypercube experimental design. The maximum stress was taken as constraints. During the optimization processes, double point adding criterion was used to update Kriging model continuously for the sake of higher global accuracy and local accuracy at the optimal solutions until obtaining optimal solutions. The research results show that the weight of hoist towers is reduced by 39.37% after optimization while the maximum stress is unchanged. The optimization efficiency is greatly improved based on the dynamic Kriging model with double point adding criterion in comparison with simulation model. The dynamic Kriging model with double point adding criterion has higher global accuracy, local accuracy and local accuracy at the optimal solutions compared with static Kriging model and dynamic Kriging model with traditional adding criterion.

Key words: dynamic Kriging model, double point adding criterion, multi-parameter constrained optimization, hoist tower, artificial bee colony algorithm

摘要： 为在满足强度要求的情况下尽量减小提升塔架质量，同时解决传统Kriging模型全局精度和局部精度不易同时保证的问题，提出了一种双加点动态Kriging模型，并利用该模型和人工蜂群算法对提升塔架进行了优化设计。将全局敏感性分析得到的敏感参数作为设计变量，利用拉丁超立方试验设计得到的样本数据建立初始Kriging模型，以最大应力为约束条件，通过人工蜂群算法对提升塔架进行了减重优化。优化过程中, 采用双加点准则不断更新Kriging模型，以提高模型的全局精度和最优解处局部精度，直到获得最优解。研究结果表明：在最大应力不变的条件下，优化后的提升塔架质量减小了39.37%。基于双加点动态Kriging模型的优化设计与仿真模型的优化设计相比，其优化效率大幅度提高。双加点动态Kriging模型相较于静态Kriging模型和基于传统加点准则的动态Kriging模型，具有更高的全局精度、局部精度和最优解处局部精度。

关键词: 动态Kriging模型, 双加点准则, 多参数带约束优化, 提升塔架, 人工蜂群算法

CLC Number:

TH213.3

CHEN Peng;ZHANG Qing;HUANG Lei. Optimization Design of Hoist Towers Based on Dynamic Kriging Model with Double Point Adding Criterion[J]. China Mechanical Engineering.

陈鹏;章青;黄磊. 基于双加点动态Kriging模型的提升塔架优化设计[J]. 中国机械工程.

References

[1]陈立娜, 张维刚. 基于Kriging模型的发动机罩多目标优化设计[J]. 中国机械工程, 2013, 24(22)：3014-3018.
CHEN Lina, ZHANG Weigang. Multi-objective Optimization Design of Engine Hood Based on Kriging Model[J]. China Mechanical Engineering, 2013, 24(22)：3014-3018.
[2]何东海. Kriging代理模型的序列优化及其应用[D]. 上海：华东理工大学, 2015.
HE Donghai. Sequential Optimization Method Based on Kriging Surrogate Model and Its Application[D]. Shanghai：East China University of Science and Technology, 2015.
[3]JONES D R, SCHONLAU M, WELCH W J. Efficient Global Optimization of Expensive Black-box Functions[J]. Journal of Global Optimization, 1998, 13:445-492.
[4]KARABOGA D. An Idea Based on Honey Bee Swarm for Numerical Optimization[R]. Kaiseri：Erciyes University, 2005.
[5]KARABOGA D, AKAY B. A Modified Artificial Bee Colony (ABC) Algorithm for Constrained Optimization Problems[J]. Applied Soft Computing, 2011, 11：3021-3031.
[6]罗钧, 林于晴, 刘学明, 等. 改进蜂群算法及其在圆度误差评定中的应用[J]. 机械工程学报, 2016, 52(16)：27-32.
LUO Jun, LIN Yuqing, LIU Xueming, et al. Research on Roundness Error Evaluation Based on the Improved Artificial Bee Colony Algorithm[J]. Journal of Mechanical Engineering, 2016, 52(16)：27-32.
[7]唐皓, 段吉安, 郑煜, 等. 平面光波导精密对准平台运动误差的敏感性分析[J]. 中国机械工程, 2012, 23(8)：888-892.
TANG Hao, DUAN Ji'an, ZHENG Yu, et al. Motion Error Sensitivity Analysis of Planar Optical Waveguide Precise Aligning Stage[J]. China Mechanical Engineering, 2012, 23(8)：888-892.
[8]SOBOL I M. On Sensitivity Estimation for Nonlinear Mathematical Models[J]. Mathmaticheskoe Modelirovanie, 1990, 2(1)：112-118.
[9]高云凯, 张朋, 吴锦妍, 等. 基于Kriging模型的大客车侧翻安全性多目标优化[J]. 同济大学学报（自然科学版）, 2012, 40(12)：1882-1887.
GAO Yunkai, ZHANG Peng, WU Jinyan, et al. Multi-objective Optimization for Bus Rollover Safety Based on Kriging Model[J]. Journal of Tongji University (Natural Science), 2012, 40(12)：1882-1887.