恶劣环境下钢包脱挂钩状态自动识别技术研究  

Research on automatic recognition technology of ladle hook and unhook state in harsh environment

作　　者：夏建超[1] 周亮亮[2] 陈仁 XIA Jian-chao;ZHOU Liang-liang;CHEN Ren(Shanghai Meishan Iron and Steel Limited by Share Ltd.,Nanjing 210039,China;Taiyuan Heavy Industry Co.,Ltd.,Taiyuan 030024,China;Beijing Wattman Intelligent Technology Co.,Ltd.,Beijing 100038,China)

机构地区：[1]上海梅山钢铁股份有限公司炼钢厂,江苏南京210039 [2]太原重工股份有限公司,山西太原030024 [3]北京瓦特曼智能科技有限公司,北京100038

出　　处：《重型机械》2023年第4期62-67,共6页Heavy Machinery

摘　　要：吊运钢包时,对挂钩和耳轴实现自动、可靠识别定位,可提高吊运过程的稳定性和安全性。采用SLAM实时定位、3D激光重建技术以及3D神经网络模型,对钢包以及起重机挂钩进行实时三维激光点云稠密建模,从而得到实时的脱挂钩状态结果。此项技术为起重机脱挂钩操作的状态确认增加了一套客观判断机制,与人工主观判断一起,进一步增加了脱挂钩状态判断的准确性与稳定性,大大提高调运钢包的作业效率,从而提高企业的生产效益。When lifting ladle,the hook and trunnion can be automatically and reliably identified and positioned,which can improve the stability and safety of the lifting process.SLAM real-time localization,3D laser reconstruction technology and 3D neural network model were used to conduct real-time 3D laser point cloud dense modeling of ladle and crane hook,so as to obtain real-time hook and unhook state results.This technology adds a set of objective judgment mechanism for the status confirmation of crane hook operation,together with manual subjective judgment,it further increases the accuracy and stability of the state judgment of hook and unhook operation,greatly improves the operation efficiency of ladle transportation,and thus improves the production efficiency of enterprises.

关 键 词：自动脱挂钩识别 SLAM 激光与图像融合 神经网络 

分 类 号：TF341.9[冶金工程—冶金机械及自动化]