摘要
为提升油气井场可燃气体泄漏巡检的精度与效率,提出一种基于机器视觉的油气井场可燃气体泄漏三维巡检方法。油气井场可燃气体泄漏三维巡检技术架构由网络通信控制模块、环境感知模块和逻辑控制模块组成,其中网络通信模块基于机器人操作系统服务,将云计算集群经无线网关与环境感知模块连接;环境感知模块集成云台式激光气体监测仪及搭载激光雷达、热成像摄像头的巡检机器人,实时采集油气井场数据;逻辑控制模块经嵌入式板规划巡检路径,驱动巡检机器人,并处理点云数据,构建PDMS三维模型,同时实现可燃气体泄漏的检测及定位。最终,工控机展示三维场景及泄漏检测结果,实现油气井场的三维巡检。实验结果表明:该方法具备较好的智能巡检机器人巡检导航规划能力、点云数据预处理能力以及可燃气体泄漏检测与定位能力,具有良好的工程应用前景。
In order to improve the inspection precision and efficiency of combustible gas leakage in oil and gas well field,a three-dimensional inspection method of combustible gas leakage in oil and gas well field based on machine vision is proposed.The network communication module connects the cloud computing cluster to the environment awareness module via the wireless gateway based on ROS service.Then the environmental perception module integrates cloud based laser gas monitoring instruments and inspection robots carried with laser radar and thermal imaging cameras to collect real-time data from oil and gas well sites.The logic control module plans the inspection path through the embedded board,drives the inspection robot,processes point cloud data,constructs a PDMS 3D model,and detects and locates combustible gas leakage.Finally,the industrial computer displays the 3D scene and leak detection results,achieving 3D inspection of oil and gas well sites.The experimental results show that this method has good patrol navigation planning ability,point cloud data preprocessing ability and combustible gas leakage detection and positioning ability of intelligent patrol robot,which has promising prospects for engineering applications.
作者
袁良
YUAN Liang(Tarim Oilfield Branch Yingmai Oil and Gas Management Zone,Korla 841000,China)
出处
《甘肃科学学报》
2025年第6期105-112,共8页
Journal of Gansu Sciences
关键词
机器视觉
油气井场
可燃气体泄漏
三维巡检
点云数据
预处理
Machine vision
Oil and gas well field
Combustible gas leakage
Three dimensional patrol inspection
Point cloud data
Pretreatment
