基于YOLOv11的新能源农机电机壳体密封性检测研究OA
Research on sealing inspection of motor housings for new energy agricultural machinery based on YOLOv11
随着新能源农机产业的快速发展,电机壳体密封性检测面临更高要求.针对传统方法存在效率低、准确性不足、无法实现实时在线检测等问题,提出了基于 YOLOv11 的视觉检测方法,通过改进骨干网络和颈部网络架构,增强特征提取能力,在提升检测精度的同时降低模型复杂度.在自建的电机壳体密封性缺陷图像数据集上的试验结果表明,YOLOv11 模型在 mAP50 指标达到 97.5％,在 mAP50-95 指标达到 77.6％,相较于 YOLOv10、YOLOv9 和 YOLOv8 模型,在精度和鲁棒性上均有显著提升.该方法为新能源农机关键部件的智能化、自动化质量检测与健康管理提供了有效的技术解决方案.
With the rapid advancement of the new energy agricultural machinery industry,motor housing seal integrity testing faces heightened demands.Addressing the limitations of traditional methods—including low efficiency,insuffi-cient accuracy,and inability to perform real-time online detection—a visual inspection approach based on YOLOv11 is proposed.By refining the backbone and neck architecture to enhance feature extraction capabilities,this method im-proves detection precision while reducing model complexity.Experimental results on a self-built image dataset of motor housing sealing defects demonstrate that the YOLOv11 model achieves 97.5％mAP50 and 77.6％mAP50-95.Compared to YOLOv10,YOLOv9,and YOLOv8 models,it exhibits significant improvements in both accuracy and robustness.This methodology provides an effective technical solution for the intelligent,automated quality inspection and health management of critical components in new energy agricultural machinery.
王普杰;黄军垒;李国文;郭书阳
河南开放大学,河南 郑州 450046||郑州信息科技职业学院,河南 郑州 450046郑州斯倍思机电有限公司,河南 郑州 450066河南省德备通信科技有限公司,河南 郑州 450001郑州航空工业管理学院,河南 郑州 450015
农业科技
YOLOv11电机壳体密封性检测视觉检测
YOLOv11motor housingseal integrity testingvisual inspection
《农机使用与维修》 2026 (5)
22-26,5
河南省科技攻关项目(252102220029,262102240061)河南省高等学校重点科研项目(26B413007)
评论