一种基于陶瓷工艺知识图谱的可推理数字化设计服务平台架构研究
Research on the Architecture of a Reasoning-Enabled Digital Design Service Platform Based on Ceramic Process Knowledge Graph
DOI: 10.12677/sea.2026.154049, PDF,   
作者: 樊尹莘:南昌师范学院文学与历史学院,江西 南昌;曹冰一, 史文雅:江苏建筑职业技术大学设计创意学院,江苏 徐州;甘赛雄, 黄建军:南昌理工学院南昌市AI无损表型测量技术与装备重点实验室,江西 南昌
关键词: 中小陶瓷企业数字化设计服务平台知识即服务工艺知识库Small and Medium-Sized Ceramic Enterprises Digital Design Service Platform Knowledge as a Service Process Knowledge Base
摘要: 近年来,中小陶瓷企业对产品设计创新能力的重视程度不断提升,但受限于设计资源获取渠道有限、工艺知识积累方式传统等因素,其设计能力和响应速度难以跟上日益增长的个性化消费需求。本文提出一种基于陶瓷工艺知识图谱的可推理数字化设计服务平台架构(简称:RDSPA),面向中小陶瓷企业提供低成本、低门槛的设计知识服务。该平台的核心是一个面向陶瓷设计领域的工艺知识图谱,通过异质图结构实现工艺知识的语义化存储,通过图注意力网络实现跨类别推理检索,并集成Beta-spline参数化建模与改进Harris角点检测,分别支撑器型连续形态表达与图像特征自动提取。模块级实验验证了改进Harris角点检测与Beta-spline曲线拟合两项核心技术的有效性,改进算法在检测精度与拟合精度上均优于经典方法。在原型图谱上开展的路径检索测试验证了推理模块在跨类别关联发现中的有效性。以工艺知识图谱和可推理服务为核心的平台化设计模式,能够有效提升中小陶瓷企业的设计能力和订单快速响应能力,是弥合其设计能力与市场需求之间差距的一条可行路径。
Abstract: In recent years, small and medium-sized ceramic enterprises have placed increasing emphasis on product design innovation capabilities. However, constrained by limited access to design resources and traditional approaches to process knowledge accumulation, their design capabilities and responsiveness have struggled to keep pace with the growing demand for personalized consumer products. This paper proposes a reasoning-enabled digital design service platform architecture based on a ceramic process knowledge graph, referred to as RDSPA, which provides low-cost, low-threshold design knowledge services for small and medium-sized ceramic enterprises. The core of the platform is a process knowledge graph oriented to the ceramic design domain, which employs a heterogeneous graph structure for semantic storage of process knowledge, utilizes a graph attention network for cross-category reasoning and retrieval, and integrates Beta-spline parametric modeling and improved Harris corner detection to support continuous form representation of vessel types and automatic image feature extraction, respectively. Modular experiments have validated the effectiveness of the two core technologies—improved Harris corner detection and Beta-spline curve fitting—demonstrating that the improved algorithms outperform classical methods in both detection accuracy and fitting precision. A path retrieval test conducted on the prototype knowledge graph has validated the effectiveness of the reasoning module in discovering cross-category associations. The platform design model, centered on the process knowledge graph and reasoning-enabled services, can effectively enhance the design capabilities and order responsiveness of small and medium-sized ceramic enterprises, offering a viable pathway to bridge the gap between their design capacity and market demands.
文章引用:樊尹莘, 曹冰一, 史文雅, 甘赛雄, 黄建军. 一种基于陶瓷工艺知识图谱的可推理数字化设计服务平台架构研究[J]. 软件工程与应用, 2026, 15(4): 533-546. https://doi.org/10.12677/sea.2026.154049

参考文献

[1] 胡开华, 黄华, 章义来, 等. 陶瓷产品三维互动装饰设计定制服务平台[J]. 重庆大学学报, 2020, 43(11): 121-136.
[2] 陈莎莉, 黄惠, 熊丹, 等. 国家试验区背景下景德镇陶瓷产业数字化转型探索[J]. 江苏陶瓷, 2024, 57(3): 3-5, 7.
[3] 罗仕鉴, 董烨楠. 面向文化创意设计的器物知识集成与管理[J]. 计算机集成制造系统, 2018, 24(4): 964-977.
[4] 王智鸿, 李锋. 用户体验设计视域下的日用陶瓷创意策略研究[J]. 陶瓷学报, 2024, 45(5): 1060-1066.
[5] 张亚林, 王静. 陶瓷产业数字化转型的路径与挑战[J]. 陶瓷学报, 2023, 44(2): 235-242.
[6] 周正达, 王昊, 汪琳, 等. ChatKG: 一种基于大语言模型和提示工程的非遗知识图谱构建框架——以中国非遗陶瓷制作工艺为例[J]. 图书馆杂志, 2026, 45(4): 82-97.
[7] 施法中. 计算机辅助几何设计与非均匀有理B样条(CAGD & NURBS) [M]. 修订版. 北京: 高等教育出版社, 2025.
[8] 武璇. 计算机辅助设计在陶瓷产品设计中的应用[J]. 陶瓷, 2024(7): 118-120.
[9] 古和今, 郑萍, 熊金泉, 等. 陶瓷产品辅助设计与可视化关键技术研究[J]. 陶瓷学报, 2018, 39(4): 475-481.
[10] 余晓凡, 甘莉, 何也, 等. 陶瓷艺术设计知识图谱构建研究[J]. 包装工程, 2022, 43(8): 247-256.
[11] 翁彦俊, 邹媛, 熊喆. 数字赋能千年瓷都——景德镇“古陶瓷基因库数字化平台”文化遗产保护实践[N]. 中国文物报, 2025-12-04(006).
[12] 汪琳, 王昊, 李晓敏, 等. 融合学习扩展的非遗陶瓷工艺领域术语库构建及应用[J]. 图书馆论坛, 2024, 44(2): 66-78.
[13] Liu, S., Mok, H.C.S., Ling, L., Klein, T. and Lc, R. (2026) ClayScape: A Genai-Supported Workflow for Designing Chinese Style Ceramics with Clay 3D Printing. Proceedings of the 2026 Designing Interactive Systems Conference, Singapore, 13-17 June 2026, 2014-2036.
https://doi.org/10.1145/3800645.3812941
[14] 肖刚等. 视觉语言模型引导的青瓷跨模态知识图谱构建[J]. 中国图象图形学报, 2025, 30(5): 1318-1333.
[15] 吕志明, 王立. 面向领域应用的需求分析方法研究[J]. 计算机科学, 2021, 48(6): 231-238.
[16] 徐梦溪, 沈克永, 涂宏斌, 王丹华. 一种视频与物流业务数据融合的效率-安全双目标协同优化系统设计模式[J]. 软件工程与应用, 2025, 14(4): 897-905.
[17] 郭磊, 赵竞, 胡钢. 基于四次ω-Bézier曲线的陶瓷产品造型设计[J]. 中国陶瓷, 2015, 51(9): 41-44.
[18] 许昕翔, 吴晟, 刘韵琦. 数字化时代背景下陶瓷艺术品生产策略研究[J]. 江苏陶瓷, 2024, 57(4): 46-48.
[19] Farin, G. (2002) Curves and Surfaces for CAGD: A Practical Guide. 5th Edition, Morgan Kaufmann.
[20] Bordes, A., Usunier, N., Garcia-Duran, A., et al. (2013) Translating Embeddings for Modeling Multi-Relational Data. Proceedings of the 27th International Conference on Neural Information Processing Systems, Lake Tahoe, 5-10 December 2013, 2787-2795.
[21] Veličković, P., Cucurull, G., Casanova, A., et al. (2018) Graph Attention Networks. arXiv: 1710.10903.
[22] Harris, C. and Stephens, M. (1988) A Combined Corner and Edge Detector. Proceedings of the 4th Alvey Vision Conference, Manchester, 31 August-2 September 1988, 147-151.
[23] Shi, J.B. and Tomasi (1994) Good Features to Track. Proceedings of IEEE Conference on Computer Vision and Pattern Recognition CVPR-94, Seattle, 21-23 June 1994, 593-600.
https://doi.org/10.1109/cvpr.1994.323794
[24] Joe, B. (1989) Multiple-Knot and Rational Cubic β-Splines. ACM Transactions on Graphics, 8, 100-120.
https://doi.org/10.1145/62054.62055