一种深立井井筒数据采集及分析系统设计

Design of data acquisition and analysis system for deep vertical shaft

目前用于深立井井筒数据采集的罐道巡检机器人能够搭载的传感器数量较少,且没有安全保护装置,存在安全隐患,此外,深立井井筒数据可视化方案大多采用3D GIS进行渲染和显示,存在不易移植、开发周期长等问题。针对上述问题,设计了一种深立井井筒数据采集及分析系统。对罐道巡检机器人进行改进,通过增加车轮锁装置保证机器人运行过程中的安全性,以带车轮锁的罐道巡检机器人作为移动平台,通过搭载红外摄像头、光电编码器、超声波测距模块及各种传感器实现深立井井筒数据采集。采用云服务器加前端可视化面板的方式进行数据处理和显示,通过云服务器接收罐道巡检机器人发送的各类数据,并对数据进行分类处理:温湿度传感器和气体传感器数据直接存储在相应文件夹内;视频数据采用卷积神经网络(CNN)进行处理,分析罐道和井壁是否出现裂缝、形变等,再将分析结果存储在相应文件夹内。对前端可视化面板进行轻量化处理,采用异步JavaScript和XML(Ajax)从云服务器定时读取数据,并采用JavaScript编写上位机界面显示程序,以提高系统的便携性和可移植性。测试结果表明:以带车轮锁的罐道巡检机器人作为数据采集装置,提高了数据采集的可靠性和安全性;可视化面板加云服务器的数据处理和显示方式将上位机软件所占内存缩小至5 MB以内,页面刷新快,可移植性强。

At present,the cage guide inspection robot used for data acquisition of deep vertical shaft can only carry a small number of sensors.There is no safety protection device,so there are potential safety hazards.In addition,most of the deep vertical shaft data visualization programs use 3D GIS for rendering and display.It is difficult to transplant and has long development cycle.In order to solve the above problems,a data acquisition and analysis system for deep vertical shaft is designed.The cage guide inspection robot is improved by adding a wheel lock device to ensure the safety of the robot during operation.The cage guide inspection robot with wheel lock is used as mobile platform.The data acquisition of deep vertical shaft is realized by infrared camera,photoelectric encoder,ultrasonic ranging module and various sensors on the robot.The cloud server plus front-end visualization panel is used for data processing and display.The cloud server receives various data sent by the cage guide inspection robot and classifies and processes the data.The data of temperature and humidity sensor and the gas sensor is directly stored in the corresponding folder.The convolutional neural network(CNN)is applied to process the video data and analyze whether there are cracks or deformation in the cage guide and the shaft wall.The analysis results are stored in the corresponding folder.The front-end visual panel is lightweighted.Asynchronous Java Script and XML(Ajax)are used to read data from the cloud server regularly.And Java Script is used to write the host computer interface display program to improve the portability of the system.The test results show that the cage guide inspection robot with wheel lock used as the data acquisition device can improve the reliability and safety of the system.The data processing and display method of the visualization panel and cloud server reduces the memory occupied by the host computer software to less than 5 MB.The page refreshes quickly and the portability is strong.