Abstract: To improve the intelligence level of the aero-engine blade assembly execution process,a method for intelligent detection and AR guidance of the aero-engine blade assembly execution process was proposed,which included three links: blade coding recognition,material kitting based on AR and status detection during the complete placement process.To solve the problem of lack of automatic identification and intelligent error correction,an aero-engine blade material management architecture based on coderecognition was built,and an image processing based on pre-processing enhancement for blade codeimages was proposed.Bayesian error correction was used to judge the recognition results and perform post-processing for error correction,which improved the recognition accuracy of blade codes.Meanwhile,in the process of manual material kitting,AR enhanced visual information was used to assist users in quickly executing task tasks,reducing the time for selecting blade materials.In addition,an error prevention and correction system based on detection and comparison was built for blade material placement,which could avoid human errors.The intelligent detection and AR guidance method proposed for aero-engine blade assembly execution process could effectively reduce the consumption of human,material resources and time,which played a technical supporting role in promoting the aero-engine towards intelligent and automatic production.

Key words: aero-engine blade, optical character recognition, post-processing, augmented reality, assembly execution process

摘要： 为了提高航空发动机叶片装配执行过程的作业智能化程度,本文提出一种航空发动机叶片装配执行过程智能检测及AR引导方法,该方法包括叶片编码识别、物料AR出入库和齐套摆放过程状态检测3个环节。针对航空发动机叶片物料缺乏自动化识别和智能化纠错的问题,搭建基于编码识别的叶片物料管理架构,提出基于图像处理的叶片编码图像前处理增强操作,并利用贝叶斯纠错对识别结果进行正误判断和纠错校正的后处理操作,提高了叶片编码识别准确率;在物料人工出入库环节,利用AR增强可视化信息辅助用户快速执行作业任务,降低了叶片物料选取作业的时间;针对叶片物料齐套准备过程,构建了基于检测比对的防错纠错系统,避免发生人为错误。所提叶片装配执行过程智能检测及AR增强辅助引导方法可以有效减少人力物力和时间消耗,在推动航空发动机迈向智能化和自动化生产上起到技术支撑作用。

关键词: 航空发动机叶片, 光学字符识别, 后处理, 增强现实, 装配执行过程