[1]侯晓辉,王金山,吴艳明,等.基于聚类与SDC算法的智能仓储系统分类识别技术研究[J].工业仪表与自动化装置,2026,(02):84-89.[doi:10.19950/j.cnki.CN61-1121/TH.2026.02.016]
　HOU Xiaohui,WANG Jinshan,WU Yanming,et al.Research on classification and recognition technology of intelligent warehouse system based on clustering and sdc algorithm[J].Industrial Instrumentation & Automation,2026,(02):84-89.[doi:10.19950/j.cnki.CN61-1121/TH.2026.02.016]
点击复制

基于聚类与SDC算法的智能仓储系统分类识别技术研究()

《工业仪表与自动化装置》[ISSN:1000-0682/CN:61-1121/TH]

卷:
期数:
2026年02期
页码:
84-89
栏目:
出版日期:
2026-04-15

文章信息/Info

Title:
Research on classification and recognition technology of intelligent warehouse system based on clustering and sdc algorithm
文章编号:
1000-0682（2026）02-0084-06
作者:
侯晓辉; 王金山; 吴艳明; 程 嵩; 齐佳丽
大唐（内蒙古
Author(s):
HOU Xiaohui;  WANG Jinshan;  WU Yanming;  CHENG Song;  QI Jiali
Datang(Inner mongolia) Energy Development Co.Ltd.,Inner Mongolia Autonomous Region Hohhot 010000, China
关键词:
智能仓储; 聚类分析; SDC; 机器视觉; 分类识别
Keywords:
intelligent warehousing;  cluster analysis;  SDC;  machine vision;  classification and recognition
分类号:
TP39
DOI:
10.19950/j.cnki.CN61-1121/TH.2026.02.016
文献标志码:
A
摘要:
该文以YOLOv8为基本框架，在主干框架中添加SDC增加图像特性提取能力，并在检测头中添加改进损失函数强化位置识别能力；使用聚类算法和神经网络耦合来输出货物信息。该文提出的改进模型在5种不同类别货物检测中，相比于无锚框算法YOLOv8，PA值分别提升了8.04%，8.26%，8.18%，7.49%和7.41%，mPA值提升了7.35%。在图像特征采集后，基于概率的聚类分析展现出良好地分类效果，对于货架层、地面层和地面货架混合层的货物进行测试。货架层和地面层的识别准确率分别为97.1%和96.8%。而混合环境下，准确率为94.2%。
Abstract:
This paper takes YOLOv8 as the basic framework, adds SDC in the main framework to enhance the ability to extract image features, and adds an improved loss function in the detection head to strengthen the position recognition ability. The clustering algorithm and the neural network coupling algorithm are used to output the goods information. The improved model proposed in this paper has increased the PA values by 8.04%, 8.26%, 8.18%, 7.49% and 7.41% respectively compared with the anchor-free algorithm YOLOv8 in the detection of five different types of goods, and the mPA value has increased by 7.35%. After image feature collection, probability-based cluster analysis demonstrated excellent classification results, and tests were conducted on goods in the shelf layer, ground layer, and mixed ground and shelf layers. The accuracy rates of goods identification on the shelf layer and the ground layer were 97.1% and 96.8% respectively. In a mixed environment, the accuracy rate was 94.2%.

参考文献/References:

[1] 李晖,刘栋,姚丹阳.面向碳达峰碳中和目标的我国电力系统发展研判[J].中国电机工程学报,2021,41(18): 6245-6259.

[2] 徐慧剑.基于物联网RFID 技术的智能仓储系统的设计与实现[J].制造业自动化,2012,34(07): 139-141.
[3] 李秉津.基于物联网技术的电力应急物资智能仓储管理模式设计[J].网络安全和信息化, 2024(09): 71-73．
[4] 林健光.基于ZigBee的智能仓储物流系统设计研究[J].装备维修技术,2023(1):38-43.
[5] 廖仕军.基于机器视觉的工业机器人定位系统研究与应用[J].无线互联科技,2021, 18(7): 63-64.
[6] 王成军,韦志文,严晨.基于机器视觉技术的分拣机器人研究综述[J].科学技术与工程,2022, 22(3): 893-902.
[7] 徐青青.基于机器视觉的工业机器人智能分拣系统设计[J].仪表技术与传感器,2019(8):92-95.
[8] 郑建彬,白雅贤,詹恩奇,等.基于改进SIFT匹配方法的货架乳制品识别[J].计算机科学,2017,44(9):315-319.
[9] Yilmazer R, Birant D. Shelf auditing based on image classification using semi-supervised deep learning to increase on-shelf availability in grocery stores [J]. Sensors, 2021, 21(2): 327.
[10] 李云娟,樊雪双.基于数学建模的PSO-DE算法在机器人智能拣货过程中的应用研究[J].自动化与仪器仪表,2024(03): 197-200.
[11] 吴海彬,林宜.基于改进Voronoi图的移动机器人在线路径规划[J]. 中国工程机械学报,2007(01): 117-121.
[12] 唐媛红,刘月云.基于PLC的水果分拣机器人动作控制优化[J].农机化研究,2021, 43(12): 233-237.
[13] 黄雨晨,张利平,唐秋华,等.带AGV开放车间调度的析取图模型与求解[J].机械设计与制造,2023(6): 10-15.

相似文献/References:

[1]何军红,郝宇航.基于AGV与RFID的智能仓储管理系统的设计与实现[J].工业仪表与自动化装置,2018,(04):42.[doi:1000-0682（2018）04-0000-00]
　HE Junhong,HAO Yuhang.Design and implementation of intelligent warehouse management system based on AGV and RFID[J].Industrial Instrumentation & Automation,2018,(02):42.[doi:1000-0682（2018）04-0000-00]
[2]高 林,李琪琪.基于交通流数据的交通状态判别算法研究[J].工业仪表与自动化装置,2020,(02):8.
　GAO Lin,LI Qiqi.Research on traffic state discrimination algorithm based on traffic flow data[J].Industrial Instrumentation & Automation,2020,(02):8.

备注/Memo

备注/Memo:
收稿日期：2025-11-25基金项目：大唐(赤峰)新能源有限公司科技项目“新能源智能仓储管理体系建设”（DTNM-2024-20340）第一作者：侯晓辉（1974—），男（蒙古族），内蒙古赤峰人，本科，高级经济师，研究方向为智能仓储。
更新日期/Last Update: 1900-01-01