工程设计学报2024，Vol.31Issue(2)：254-262,9.DOI:10.3785/j.issn.1006-754X.2024.03.147
基于树莓派和视觉图像的钻井振动筛倾角调节系统
Inclination angle adjustment system for shale shaker based on Raspberry Pi and visual image
摘要
Abstract
At present,the shale shakers in China are operated by manually detecting the solid-liquid separation status and manually adjusting the inclination angle of the screen surface,which can not realize self-adaptive work,and the phenomenon of drilling fluid"running"often occurs.To solve this problem,an inclination angle adjustment system for shale shaker based on Raspberry Pi and visual image was proposed.The system used Raspberry Pi,dedicated camera,motor drive board and two stepper motors as hardware platform,and was equipped with an image recognition software developed based on OpenCV and improved AlexNet model,which could achieve visual detection of screen surface solid-liquid separation status and automatic adjustment of screen surface inclination angle of the shale shaker.Firstly,according to the position characteristics of the liquid phase termination line during the solid-liquid separation process of shale shaker,the collected screen surface images were divided into three categories:normal,low mud and slurry running state,and the screen surface image dataset was constructed.Then,the AlexNet model based on transfer learning was constructed using TensorFlow platform to automatically recognize the screen surface solid-liquid separation status of shale shaker.Finally,based on the recognition results,two stepper motors were controlled synchronously by the GPIO(general purpose input/output)interface of Raspberry Pi to realize the inclination angle adjustment for the shale shaker.The results showed that the accuracy of the designed inclination angle adjustment system for recognizing screen surface solid-liquid separation status reached 97.3%,and the response time was about 1.5 s,which could meet the inclination angle adjustment requirements of the shale shaker.The inclination angle adjustment system equipment has small volume,low cost and is easy to debug and maintain,which can effectively improve the automation level of shale shakers.关键词
钻井振动筛/树莓派/倾角调节系统/AlexNet模型/图像识别Key words
shale shaker/Raspberry Pi/inclination adjustment system/AlexNet model/image recognition分类
石油、天然气工程引用本文复制引用
侯勇俊,贾文俊,刘博文,吴先进..基于树莓派和视觉图像的钻井振动筛倾角调节系统[J].工程设计学报,2024,31(2):254-262,9.基金项目
四川省科技计划项目(2021YFG0261,2022YFQ0064) （2021YFG0261,2022YFQ0064）
