CASIA OpenIR  > 毕业生  > 硕士学位论文
Alternative TitleDevelopment and Research of Fault Diagnosis Expert System for CNC Machine Tool
Thesis Advisor赵晓光
Degree Grantor中国科学院研究生院
Place of Conferral中国科学院自动化研究所
Degree Discipline控制理论与控制工程
Keyword数控机床 故障诊断 专家系统 故障树 产生式表示 小波包算法 Bp神经网络 正反向推理 Cnc Machine Tool Fault Diagnosis Expert System Fault Tree Production Rule Wavelet Packet Algorithm Bp Neural Network Forward And Backward Reasoning
Abstract作为机械制造业的基础加工装备,数控机床在我国的国民经济发展中发挥着重要作用,数控机床故障的发生使得其精度和可靠性降低,影响加工工件的质量,严重时还会导致停机或报废,使企业和国家蒙受巨大的经济损失。因此,开展数控机床故障诊断技术研究具有重大的理论和现实意义。 本文是在国家“十一五”“高档数控机床与基础制造装备”科技重大专项课题“总线式全数字高档数控装置:智能故障诊断技术(2009ZX040009-013)”的支持下完成的,主要对数控机床故障诊断专家系统进行了研究,开发了通用的数控机床状态监测和故障诊断专家系统。 首先,对数控机床故障诊断专家系统的整体架构进行了设计,针对传统专家系统通用性不强的问题,在对故障树分析法进行研究的基础之上,提出了一种基于故障树分析法的开放式建模方式,使得用户能够根据不同型号机床的具体情况交互式地建立故障树,并采用产生式规则的表示形式对知识进行存储,为后续故障诊断提供依据。 其次,为了能够及时准确地提取故障征兆,提出了利用传感器信息进行辅助故障征兆提取的方法。针对实时采集到的噪声和振动非平稳信号,利用小波包算法进行信号特征分析并通过改进的BP神经网络进行状态分类,实现了故障征兆提取的目标,采用融合多种传感器信息的正反向混合推理策略,实现了对故障原因的精确定位。 最后,基于Visual C++开发工具和ADO数据库连接技术开发了通用的数控机床状态监测和故障诊断专家系统软件,并在自行搭建的实验平台上进行了静态库诊断、交互式诊断以及基于多传感器信息的在线诊断模拟实验,验证了故障诊断专家系统的可行性和有效性。
Other AbstractAs the fundamental processing equipment of the mechanical manufacturing industry, CNC machine tools play an important role in China's national economic development. Once failure happens, the precision and reliability of CNC machine tools will degrade, which will affect the quality of the work piece. In severe cases, the failure will result in machine shutdown or scrap, which will bring huge economic loss to enterprises and countries. Therefore, conducting research on the fault diagnosis technology of CNC machine tools has great theoretical and practical significance. This dissertation is supported by the project Digital High-end NC Device with Bus Organization: Intelligent Fault Diagnosis Technology (2009ZX040009-013) of the National Science and Technology Major Project High-end CNC Machine tools and Fundamental Manufacturing Equipment of the Eleventh Five-Year Plan. The research on CNC machine tool fault diagnosis expert system is conducted and a general condition monitoring and fault diagnosis expert system for CNC machine tool is developed. Firstly, the overall structure of the CNC machine tool fault diagnosis expert system is designed. The versatility of the traditional expert system is limited, to solve this problem, an open modeling method based on fault tree analysis is proposed, which allows the user to build fault trees for different types of CNC machine tools in an interactive way. The information of the fault trees is then stored into database in the form of production rule, serving as the basis for subsequent fault diagnosis. Secondly, in order to extract the fault symptoms timely and accurately, an auxiliary fault symptom extraction method using the sensor information is proposed. For the non-stationary noise and vibration signals acquired in real time, wavelet packet algorithm is adopted for signature analysis and the improved BP neural network method is employed for state classification, which helps to achieve the goal of fault symptom extraction. Precise positioning of the cause of malfunction is realized by means of forward and backward reasoning with the integration of multi-sensor information. Finally, a general condition monitoring and fault diagnosis expert system software for CNC machine tool based on the Visual C++ development tool and ADO database connectivity technology is developed. For the designed expert system, many simulation experiments are conducted on the experimental platform, such as the static library diagnosis, the ...
Other Identifier200928014628022
Document Type学位论文
Recommended Citation
GB/T 7714
张爱瑜. 数控机床故障诊断专家系统开发与研究[D]. 中国科学院自动化研究所. 中国科学院研究生院,2012.
Files in This Item:
File Name/Size DocType Version Access License
CASIA_20092801462802(6298KB) 暂不开放CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[张爱瑜]'s Articles
Baidu academic
Similar articles in Baidu academic
[张爱瑜]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[张爱瑜]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.