CNC刀具工况集成视诊系统

发布时间：2018-03-01 09:55

本文关键词： CNC 刀具工况视诊系统集成　出处：《南昌大学》2017年硕士论文　论文类型：学位论文

【摘要】：面对第四次工业革命,为推进我国从制造业大国向制造业强国的历史性转变,《中国制造2025》国家战略对制造加工过程自动化、智能化提出了更高要求。切削过程刀具工况对产品质量和生产安全至关重要,研发一套CNC(数控机床)刀具工况集成视诊系统,对提高智能制造产业水平具有重要意义。以车削加工过程为研究对象,分析刀具磨损机理和形态,研究了基于刀具主后刀面和刀尖轮廓的视诊原理,包括图像预处理方法,颜色、纹理和形状的特征提取方法,以及基于momentum梯度下降法的DNN刀具工况分类模型。针对SINUMERIK 840D数控系统,设计了一套CNC刀具工况集成视诊系统。从嵌入式计算机系统、图像采集装置、照明装置和CNC通信模块4方面对硬件体系进行电路设计,并完成硬件实现与测试。按照下位机软件、PROFIBUS-DP通信软件、图像处理软件和数控系统集成软件4部分对软件体系进行研发,并完成代码编制和调试,包括实现了嵌入式操作系统的移植和烧写、驱动编写和应用软件开发,完成了PROFIBUS-DP从站程序开发,完成了云主机图像处理软件的开发。通过在840D数控系统上开发视诊功能M指令、完成M代码的解码及PROFIBUS-DP主站软件的编写,实现了视诊系统与数控系统的集成。通过实验对比分析不同拓扑结构和算法的刀具工况辨识模型,验证了本文设计的DNN模型的准确性和高效性。在HTM125600车铣复合加工中心进行了实验,视诊系统以功能模块形式嵌入CNC的数控系统中,实验结果表明:通过调用集成于数控系统的视诊功能子程序成功实现刀具图片的采集、图像预处理、特征提取和工况识别,证明了视诊系统具备自动化和智能化的刀具工况监测效果。
[Abstract]:In the face of the 4th Industrial Revolution, in order to promote the historic transformation from a large manufacturing country to a strong manufacturing country, the "made in China" national strategy automates the manufacturing process. Intelligent tool condition is very important to product quality and production safety. A CNC (NC machine tool) tool condition integrated inspection system is developed. It is of great significance to improve the level of intelligent manufacturing industry. Taking turning process as the research object, this paper analyzes the mechanism and shape of tool wear, and studies the principle of visual diagnosis based on the main and rear cutter face and the contour of cutter tip, including the method of image preprocessing. The feature extraction method of color, texture and shape, as well as the classification model of DNN cutting tool condition based on momentum gradient descent method. Aiming at SINUMERIK 840D numerical control system, a set of CNC tool operating condition integrated visual diagnosis system is designed. Four aspects of image acquisition device, lighting device and CNC communication module are designed, and hardware realization and test are completed. According to the software of lower computer, PROFIBUS-DP communication software, Four parts of image processing software and NC system integration software research and development of software system, and complete the code preparation and debugging, including the implementation of embedded operating system transplantation and burning, driver writing and application software development, The PROFIBUS-DP slave station program is developed, and the image processing software of cloud host is developed. By developing the M instruction of visual diagnosis function on 840D NC system, the decoding of M code and the compiling of PROFIBUS-DP master station software are accomplished. The integration of visual diagnosis system and numerical control system is realized, and the tool condition identification model of different topologies and algorithms is compared and analyzed through experiments. The accuracy and efficiency of the DNN model designed in this paper are verified. The experiment is carried out in the HTM125600 turn-milling compound machining center. The inspection and diagnosis system is embedded in the NC system of CNC in the form of function module. The experimental results show that the tool image collection, image preprocessing, feature extraction and condition recognition are successfully realized by calling the subprogram of the inspection function integrated in the NC system. It is proved that the inspection system has automatic and intelligent tool condition monitoring effect.
【学位授予单位】：南昌大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TG659

【相似文献】