石油输送场站巡检机器人设计
Patrol Inspection Robot Design for Oil Transportation Station

石油化工企业的输油场站承载着成品油的储存、运输及终端销售任务，输油场站的巡检工作十分重要，目前主要采用人工巡检的方式，人工方式危险性高、劳动强度大，而且受巡检工个人工作能力的限制，巡检质量也参差不齐。为保证输油场站的安全，研制了一种输油场站巡检机器人系统，实现24小时不间断巡检。本文阐述了巡检机器人系统的组成和实现方法，介绍了基于图像处理技术的输油场站仪表读数识别研究，根据指针式仪表的特征，对读数进行识别，识别精度满足场站巡检抄表要求；并且对输油场站巡检机器人进行路径规划研究，针对机器人在静态障碍物环境中随机出现动态障碍物情况，提出一种改进A*算法与改进人工势场法结合的混合算法，成功解决遇到动态障碍物无法避障的问题。应用结果表明，该巡检机器人系统工作可靠，取得了较好的效果。
The oil transportation stations of petrochemical enterprises are responsible for the storage, trans-portation, and terminal sales of finished oil. The patrol inspection work of oil transportation stations is very important. Currently, manual patrol inspection is mainly used, which is highly hazardous and labor intensive. Moreover, due to the limitations of the personal work ability of patrol inspec-tors, the quality of patrol inspection is also uneven. In order to ensure the safety of oil transportation stations, a patrol robot system for oil transportation stations is developed, which realizes 24-hour uninterrupted patrol inspection. This paper describes the composition and implementation method of the patrol robot system, and introduces the research on the instrument reading recognition of oil transportation stations based on image processing technology. According to the characteristics of pointer-type instruments, the readings are recognized, and the recognition accuracy meets the requirements of station patrol meter reading; in addition, a path planning study was conducted on the inspection robot for oil transportation stations. Aiming at the situation where the robot randomly encounters dynamic obstacles in a static obstacle environment, a hybrid algorithm combining the improved A* algorithm and the improved artificial potential field method was proposed, successfully solving the problem of encountering dynamic obstacles that cannot be avoided. The application results show that the patrol robot system works reliably and achieves good results.