A Hybrid Prediction Method for Beam to Upright Connectors in Industrial Racks Based on Physical Simulation and Convolutional Neural Networks

KE Simin; WANG Xinyu; GUO Chunfang; GUO Wensong; LYU Zhijun

doi:10.19554/j.cnki.1001-3563.2026.03.012

PDF(1572 KB)

Packaging Engineering ›› 2026, Vol. 47 ›› Issue (3) : 110-118. DOI: 10.19554/j.cnki.1001-3563.2026.03.012

A Hybrid Prediction Method for Beam to Upright Connectors in Industrial Racks Based on Physical Simulation and Convolutional Neural Networks

KE Simin¹, WANG Xinyu¹, GUO Chunfang¹, GUO Wensong², LYU Zhijun^1,2,*

Author information +

History +

Abstract

The work aims to propose a hybrid prediction method that integrates computational accuracy with efficiency to address the challenges of applying traditional finite element methods (FEM), which involves complicated modeling and low computational efficiency, in evaluating the performance of beam to upright connectors in industrial racks. Firstly, a finite element model was established according to the actual stress characteristics of beam to column connectors, and the effects of structural parameters on connector stiffness and failure moment were systematically examined through an orthogonal experimental design, thereby generating a simulation database. On this basis, a convolutional neural network was introduced to develop a performance prediction model for beam to upright connectors. The research results indicated that the connector stiffness was significantly affected by the number of rivets, whereas the failure moment was more strongly affected by the height of the beam. The established neural network prediction model was shown to be consistent with the finite element analysis results in terms of accuracy, while computational efficiency was improved by approximately 45 times. The work provides an effective engineering technique for the rapid and accurate assessment of beam to upright connector performance.

Key words

beam to upright connector / physical simulation / finite element method (FEM) / convolutional neural networks (CNN) / orthogonal experimental design

Cite this article

EndNote

Ris (Procite)

Bibtex

Download Citations

KE Simin, WANG Xinyu, GUO Chunfang, GUO Wensong, LYU Zhijun. A Hybrid Prediction Method for Beam to Upright Connectors in Industrial Racks Based on Physical Simulation and Convolutional Neural Networks[J]. Packaging Engineering. 2026, 47(3): 110-118 https://doi.org/10.19554/j.cnki.1001-3563.2026.03.012

References

[1] 李勇昭. 2024年中国仓储业市场发展回顾与2025年展望[J]. 中国储运, 2025(4): 4-6.
LI Y Z. A Review of the2024 Chinese Warehousing Industry Market and Outlook for 2025[J]. China Storage & Transport, 2025(4): 4-6.
[2] 贾县民, 屈亚美. 低碳物流国内研究综述[J]. 包装工程, 2022, 43(15): 289-300.
JIA X M, QU Y M.Review of Domestic Research on Low-Carbon Logistics[J]. Packaging Engineering, 2022, 43(15): 289-300.
[3] 黄兆祺. 高位立体库钢货架力学行为与稳定设计方法研究[D]. 上海: 同济大学, 2022.
HUANG Z Q.Structural Behavior and Stability Design Method of High-Rise Asymmetrically Braced Steel Storage Racks[D]. Shanghai: Tongji University, 2022.
[4] Standards Australia/Standards New Zealand. Steel Storage Racking: AS/NZS 4084[S]. Sydney: Standards Australia/Standards New Zealand, 2012.
[5] European Committee for Standardization (CEN). Steel Static Storage Systems—Adjustable Pallet Racking Systems—Principles for Structural Design: EN 15512[S]. Brussels: European Committee for Standardization, 2009.
[6] Rack Manufacturers Institute.Specification for the Design, Testing and Utilization of Industrial Steel Storage Racks: RMI[S]. Charlotte: Rack Manufacturers Institute, 2012.
[7] 国家市场监督管理总局, 国家标准化管理委员会. 立体仓库货架系统设计规范: GB/T 39681—2020[S]. 北京: 中国标准出版社, 2020.
State Administration for Market Regulation, Standardization Administration of the People's Republic of China. Racking Design Code for Steel Static Storage Systems: GB/T 39681-2020[S]. Beijing: Standards Press of China, 2020.
[8] MOHAMAD ROSLI N A, TAN C G, HASHIM H, et al. Experimental Investigation and Parameter Analysis of Boltless Steel Pallet Rack Beam-to-Column Connection[J]. Journal of Constructional Steel Research, 2025, 231: 109578.
[9] YAZICI C, ARIK E, ÖZKAL F M.Influence of Design Parameters on Beam-to-Column Connections of Steel Storage Rack Systems[J]. Engineering Failure Analysis, 2023, 152: 107439.
[10] 黄晨宇, 李陈民军, 黄曦, 等. 低周反复加载作用下工业货架梁柱节点性能退化的数值分析[J]. 计算机辅助工程, 2022, 31(1): 60-64.
HUANG C Y, LI C, HUANG X, et al.Numerical Analysis of Performance Degradation of Industrial Rack Beam-Column Joint under Low-Cycle Reversed Loading[J]. Computer Aided Engineering, 2022, 31(1): 60-64.
[11] CHENG B, WU Z Y.Investigation of Shear Stiffness of Spine Bracing Systems in Selective Pallet Racks[J]. The Open Construction and Building Technology Journal, 2015, 9(1): 1-6.
[12] MISHRA S, AGGARWAL S, SUTARIA B M.Performance Prediction of Journal Bearings with Combination of Grooves of Different Cross-Sectional Shapes through a Machine Learning Approach[J]. Engineering Research Express, 2025, 7(2): 025572.
[13] 孙晨静, 冯茹苹, 杨渊元, 等. 基于深度学习和功能磁共振影像的脑网络检测研究综述[J]. 陕西师范大学学报(自然科学版), 2025, 53(6): 15-31.
SUN C J, FENG R P, YANG Y Y, et al.Brain Network Detection and Its Application Based on Deep Learning and Functional Magnetic Resonance Imaging[J]. Journal of Shaanxi Normal University (Natural Science Edition), 2025, 53(6): 15-31.
[14] LYU Z J, CHEN Q, JI M H, et al.Physics and Data Driven Modeling for Local Damage Evolution of Cold-Formed Steel Rack Uprights: Cases Study[J]. International Journal of Steel Structures, 2025, 25(3): 755-772.
[15] 孟祥泽. 基于深度卷积神经网络的图像目标检测算法现状研究综述[J]. 数字技术与应用, 2021, 39(1): 112-116.
MENG X Z.A Survey of Current Research on Image Target Detection Algorithms Based on Deep Convolutional Neural Networks[J]. Digital Technology & Application, 2021, 39(1): 112-116.
[16] Schaefer System International Sdn. Bhd. Bending Tests on Beam End Connector According to EN 15512:2009/A.2.4 and AS 4084:2012 Section 7.5.1 of 1600 Pallet Racking System for KS TB1004015 to 1600 Upright[R]. Selangor: Malaysian Iron and Steel Industry Federation, 2021.
[17] The SPSSAU project (2025). SPSSAU. (Version 25.0) [Online Application Software][EB/OL]. https://www.spssau.com.
[18] 颜虹, 徐勇勇. 医学统计学[M]. 3版. 北京: 人民卫生出版社, 2015.
YAN H, XU Y Y.Medical Statistics[M]. 3rd ed. Beijing: People's Medical Publishing House, 2015.
[19] 杨寒雨, 赵晓永, 王磊. 数据归一化方法综述[J]. 计算机工程与应用, 2023, 59(3): 13-22.
YANG H Y, ZHAO X Y, WANG L.Review of Data Normalization Methods[J]. Computer Engineering and Applications, 2023, 59(3): 13-22.
[20] 李亚鸣, 邢凯, 邓洪武, 等. 基于小样本无梯度学习的卷积结构预训练模型性能优化方法[J]. 计算机应用, 2022, 42(2): 365-374.
LI Y M, XING K, DENG H W, et al.Derivative-Free Few-Shot Learning Based Performance Optimization Method of Pre-Trained Models with Convolution Structure[J]. Journal of Computer Applications, 2022, 42(2): 365-374.