摘要
针对传统笼绞机线轴人工巡视效率低、视角受限、安全风险高等问题，设计并实现了一种基于机器视觉的智能监测与自动化控制系统。系统采用高帧率工业相机配合环形光源进行图像采集，基于RK3588边缘计算平台部署优化后的YOLOv11目标检测算法，实现对线轴有序缠绕、无序缠绕、夹线、断线等多种状态的实时检测。通过模型压缩和INT8量化技术，系统在保证检测精度的同时实现了边缘端的实时推理。测试结果表明，系统对各类状态的检测准确率均超过92%，其中断线检测准确率高达98.2%。该系统能够实现24小时无人值守监控，有效提高了生产效率，降低了安全风险，为工业自动化生产提供了可靠的技术支撑。
关键词： 机器视觉；笼绞机；智能监测；边缘计算
Abstract
To address the problems of low efficiency, limited viewing angle, and high safety risks in traditional manual inspection of cage stranding machine spools, this paper designs and implements an intelligent monitoring and automatic control system based on machine vision. The system uses high frame rate industrial cameras with ring light sources for image acquisition, and deploys optimized YOLOv11 object detection algorithm on RK3588 edge computing platform to achieve real-time detection of various spool states including orderly winding, disorderly winding, wire clamping, and wire breaking. Through model compression and INT8 quantization techniques, the system achieves real-time inference at the edge while maintaining detection accuracy. Test results show that the detection accuracy for various states exceeds 92%, with wire breaking detection accuracy reaching 98.2%. The system can achieve 24-hour unmanned monitoring, effectively improving production efficiency and reducing safety risks, providing reliable technical support for industrial automated production.
Key words： Machine vision; Cage stranding machine; Intelligent monitoring; Edge computing
参考文献 References
[1] 许志文,廉迎战,陈阳,等.基于机器视觉的纺织品缺陷检测系统的研究与实现[J].工业控制计算机,2014,27(11): 20-21.
[2] 晋博,蔡念,夏皓,等.基于深度学习的工业视觉检测系统[J].计算机工程与应用,2019,55(02):266-270.
[3] 张晨,高盛华,王朔,等.基于计算机视觉的化工安监系统[J].计算机应用与软件,2025,42(05):72-77.
[4] 吴兆中,张新娜,王栋,等.基于机器视觉的凸轮轴表面缺陷检测实验研究[J].机电工程,2025,42(04):697-705.
[5] 付山,谢晖,易建业.基于机器视觉的金属表面缺陷检测[J].电镀与精饰,2024,46(12):145-146.
[6] 郭庆梅,刘宁波,王中训,等.基于深度学习的目标检测算法综述[J].探测与控制学报,2023,45(06):10-20.
[7] Khanam R, Hussain M. Yolov11: An overview of the key architectural enhancements[J]. arXiv preprint arXiv:2410. 17725, 2024.