An effective retrieval method for 3D models in plastic injection molding for process reuse

被引:6
作者
Guo, Fei [1 ]
Liu, Jiahuan [1 ]
Zhou, Xiaowei [1 ]
Wang, Hui [2 ]
Zhang, Yun [1 ]
Li, Dequn [1 ]
Zhou, Huamin [1 ]
机构
[1] Huazhong Univ Sci & Technol, Sch Mat Sci & Engn, State Key Lab Mat Proc & & Mould Technol, Wuhan 430074, Peoples R China
[2] Wuhan Univ Technol, Hubei Key Lab Adv Technol Automot Components, Wuhan 430070, Peoples R China
基金
中国国家自然科学基金;
关键词
Intelligent manufacturing; Injection molding; 3D model retrieval; Autoencoder; OBJECT RECOGNITION;
D O I
10.1016/j.asoc.2020.107034
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
The wide application of 3D models in computer-aided engineering (CAE) has created an urgent need for 3D model retrieval systems in manufacturing. However, recent methods are mainly based on the shape similarity of models, limiting reuse of the manufacturing process information associated with retrieved models. In this paper, we present a novel 3D model retrieval method for plastic injection molding involving process-related features. An effective feature is proposed to characterize both the geometry and process information of the 3D model, using the pressure profile based on the molding process. A variational autoencoder (VAE) is utilized to refine process-related features through unsupervised learning to improve retrieval efficiency. A 3D model database containing 120 models in actual production was built for validation experiments. The experimental results show that the proposed encoded pressure feature outperforms conventional methods with an accuracy of 86.61% compared to 78.57% using shape distribution and 73.21% using numerical features. A retrieval application proves that the information of retrieved models can be reused by a new product through the proposed method. There is considerable potential for utilizing the proposed method in similar manufacturing fields. (C) 2020 Elsevier B.V. All rights reserved.
引用
收藏
页数:12
相关论文
共 40 条
  • [1] Alexander M. B., 2008, NUMERICAL GEOMETRY N, DOI [10.1017/CBO9781107415324.004, DOI 10.1017/CBO9781107415324.004]
  • [2] The Algorithm Development and Implementation for 3D Printers based on Adaptive PID Controller
    Altan, Aytac
    Hacioglu, Rifat
    [J]. JOURNAL OF POLYTECHNIC-POLITEKNIK DERGISI, 2018, 21 (03): : 559 - 564
  • [3] [Anonymous], 2005, P 2005 ACM S SOLID P, DOI DOI 10.1145/1060244.1060256
  • [4] Deep Common Semantic Space Embedding for Sketch-Based 3D Model Retrieval
    Bai, Jing
    Wang, Mengjie
    Kong, Dexin
    [J]. ENTROPY, 2019, 21 (04)
  • [5] 3D Shape Matching via Two Layer Coding
    Bai, Xiang
    Bai, Song
    Zhu, Zhuotun
    Latecki, Longin Jan
    [J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2015, 37 (12) : 2361 - 2373
  • [6] Berg AC, 2005, PROC CVPR IEEE, P26
  • [7] Automated retrieval of 3D CAD model objects in construction range images
    Bosche, F.
    Haas, C. T.
    [J]. AUTOMATION IN CONSTRUCTION, 2008, 17 (04) : 499 - 512
  • [8] COHEN JD, 2005, VISUALIZATION HDB, P393
  • [9] Fang R, 2008, LECT NOTES COMPUT SC, V5358, P381, DOI 10.1007/978-3-540-89639-5_37
  • [10] Fei-Fei L, 2005, PROC CVPR IEEE, P524