雨污管道巡检机器人嵌入式系统的设计与实现

Design and implementation of embedded system for rain and sewage pipeline inspection robot

城市雨污排水管道的巡检主要以人工为主,但由于地下管道环境错综复杂并存在有毒可燃气体,人员下井存在很大危险.虽然有部分巡检工作尝试采用巡检机器人代替人工,但这些机器人采用的有缆方式,存在巡检距离受限、操作繁琐、能耗高等问题,导致巡检效率低下,智能化程度低.针对上述问题以及巡检机器人的任务需求,设计并实现了一种更加安全可靠与智能的雨污管道巡检机器人嵌入式系统.该设计基于CCTV巡检方式,采用STM32作为主控芯片,搭载工业摄像头、传感器组等模块采集数据,并采用无线通信的方式与上位机进行连接,之后通过YOLOV5s自主判断所采集的视频流中是否存在淤积物堵塞问题.通过在广州市番禺区的实地试验,结果表明,嵌入式系统设计合理,管道巡检机器人能够实现各项功能与技术要求.

At present,the patrol inspection of urban rainwater and sewage drainage pipelines was mainly manual.However,due to the complex environment of underground pipelines and the presence of toxic and combustible gases,there was a great danger for personnel going down the well.Although some patrolling robots tried to replace manual work,the cabled way adopted by these robots had problems such as limited patrolling distance,cumbersome operation,and high energy consumption,resulting in low patrolling efficiency and low intelligence.Aiming at the above problems and the task requirements of the inspection robot,this paper designed and implemented a more secure,reliable,and intelligent embedded system for the inspection robot of rain and sewage pipelines.This design was based on CCTV patrol mode.STM32 was used as the main control chip,equipped with industrial cameras,sensor groups,and other modules to collect data,and connected with the upper computer through wireless communication.After that,YOLOV5s could independently determine whether there was sediment blockage in the collected video stream.Through the field test in Panyu District,Guangzhou,the results showed that the embedded system design was reasonable,and the pipeline inspection robot could achieve various functions and technical requirements.