首页|用于制动梁检修的工业3D视觉检测系统设计

用于制动梁检修的工业3D视觉检测系统设计

扫码查看
因制动梁形状不规则不便机具装夹,完全靠手工检修耗时耗力,为此设计用于制动梁检修的工业3D视觉检测系统.分析3D视觉系统检测原理,结合PC端数据存储以及故障检测功能,设计制动梁的新型3D视觉检测系统,在ASODVS(主动式3D视觉传感器)的实焦点和虚焦点连线中点处增添反光片,令成像位于实焦点处,获取制动梁几何信息,并选择红外滤光片作为二次反射镜,由ODVS(3D视觉传感器)获取制动梁纹理和色彩信息;将采集到的制动梁图像信息作为PC端卷积神经网络模型输入,输出制动梁检测结果,并通过页面展现详细检修结果.仿真结果表明:该方法采集制动梁图像质量高,使制动梁细节得以全方面呈现;能够精确地识别和定位制动梁故障,完成制动梁各项参数的检测;检测结果准确可靠,并支持制动梁检修车辆信息查询.
Design of industrial 3D visual inspection system for brake beam repair
Due to the irregular shape of the brake beam,it is inconvenient for the equipment to be clamped,and relying solely on manual maintenance is time-consuming and labor-intensive.Therefore,an industrial 3D visual inspection system for brake beam ma-intenance is designed.Analyze the detection principle of the 3D vision system,combine PC data storage and fault detection functions,design a new 3D vision detection system for the brake beam,add reflective sheets at the midpoint of the line connecting the real and virtual focal points of the ASODVS(active 3D vision sensor),place the imaging at the real focal point,obtain the geometric informa-tion of the brake beam,and select an infrared filter as the secondary reflector,Obtaining texture and color information of the brake beam using ODVS(3D visual sensor);Take the collected image information of the brake beam as the input of the PC convolutional neural network model,output the brake beam detection results,and display the detailed maintenance results through the page.The Simulation results show that this method has high quality image acquisition of the brake beam,enabling comprehensive presentation of the details of the brake beam;Be able to accurately identify and locate brake beam faults,and complete the detection of various pa-rameters of the brake beam;The detection results are accurate and reliable,and support the query of vehicle information for brake beam maintenance.

brake beam3D visiondetection systemASODVSimage acquisitionconvolutional neural network

栾锋、赵军、刘宇、杨胜、王印辉

展开 >

国能铁路装备有限责任公司沧州机车车辆维修分公司,河北沧州 061100

中铁工程设计咨询集团有限公司,北京 100055

北京铁科合力科技有限责任公司,北京 100082

制动梁 3D视觉 检测系统 ASODVS 图像采集 卷积神经网络

2024

自动化与仪器仪表
重庆工业自动化仪表研究所,重庆市自动化与仪器仪表学会

自动化与仪器仪表

CSTPCD
影响因子:0.327
ISSN:1001-9227
年,卷(期):2024.(5)