摘要
为了满足目前电铲铲齿断裂的检测要求,基于深度学习的目标检测技术设计了铲齿脱落检测系统,搭建硬件平台并采集铲齿的关键帧图像。最后通过现场试验测试该系统的功能,试验结果表明,该系统能有效识别铲齿脱落状态,报警准确率可达到90%以上,对环境鲁棒性强,具有较高的可信度。
In order to meet the current detection requirements of shovel tooth fracture,designed a shovel t ooth falling off detection system based on the target detection technology in deep learning,built a hardware platform and collected the key frame images of shovel teeth.Finally,the function of this system was tested through field test.The test results show that this system can effectively identify the shovel tooth falling off state,the alarm accurate rate can reach more than 90%,has strong robustness to the environment and high reliability.
作者
缪卫峰
李雪健
杜勇志
卢进南
Miao Weifeng;Li Xuejian;Du Yongzhi;Lu Jinnan(Shenhua Baorixile Energy Co.,Ltd.,Hulunbei'er 021025,China;College of Mechanical Engineering,Liaoning Technical University,Fuxin 123000,China)
出处
《煤矿机械》
2022年第3期179-182,共4页
Coal Mine Machinery
关键词
铲齿断裂
关键帧图像
目标检测
环境鲁棒性强
broken shovel tooth
key frame image
target detection
strong environmental robustness