摘要
在多品种小批量离散制造过程中有工艺图样、车间面板记录、机床传感器信号和质检报告单等多源异构工艺信息,如何高效地获取这些来源不一、结构多样的数据,并有效实现数据可视化、挖掘数据的潜在价值是这一类相关制造型企业面临的一个共性问题.传统纸质化工艺信息管理过程中存在信息化程度低、共享性差、获取不实时和安全性弱等弊端.为解决这一问题,研究以齿轮这一典型多品种小批量离散加工产品为对象,提出一种可综合运用图像识别、文字提取、二维码技术和TCP/IP通信等技术的多元异构数据集成方法,设计并开发了可支持文本、图像和传感器数据等多源异构数据集成的智能运维系统,实现了齿轮关键工序的多源异构数据的实时集成获取、轻量化高效管理、价值挖掘及可视化应用.
Abstract
In the multi-variety small-batch discrete manufacturing process,there are multi-source heterogeneous process informa-tion such as process drawings,workshop panel records,machine tool sensor signals,quality inspection reports,etc.How to effi-ciently obtain these data from different sources and diverse structures,effectively realize data visualization,and tap the potential value of data is a common problem faced by this kind of enterprises.In the process of traditional paper-based process information management,there are some disadvantages,such as low informatization degree,poor sharing,not real-time acquisition and weak security.In order to solve this problem,it takes gear,a typical multi-variety and small-batch discrete processing product,as the object,proposes a multi-heterogeneous data integration method that can comprehensively use image recognition,text extraction,two-dimensional code technology,TCP/IP communication and other technologies,and designs and develops an intelligent opera-tion and maintenance system that can support the integration of multi-source heterogeneous data such as text,image and sensor data.It realizes real-time integration of multi-source heterogeneous data acquisition,lightweight and efficient management,value mining and visual application of key gear processes.
基金项目
广东省科技创新战略专项项目(210907104531267)
维普
万方数据