面向人机协同诊断的数控机床故障知识演化技术研究

发布时间：2018-05-03 15:52

本文选题：故障诊断 + 人机协同　；参考：《重庆大学》2015年博士论文

【摘要】：数控机床是一个集成了机械、电气、液压、气压等为一体的复杂系统,其故障的发生具有多源性、关联性、突发性、随机性等特点,只依靠传感装置对其进行数据采集根本无法满足存在大量关联特性和监测盲区的数控机床故障诊断的需要,因此,数控机床故障诊断不可避免的需要人的参与。对于分布区域广泛的数控机床而言,其绝大多数故障知识都是来源于机床设计制造企业和使用用户,因此,故障诊断经验知识具有分散性特点。而在具体某台数控机床故障诊断过程中,故障特征分析又必须依赖于相关理论及其它诊断历史经验作为参考。因此,集成人机协同知识、理论知识和维修维护经验知识对数控机床开展故障诊断意义重大。特别地,故障诊断知识在使用过程中,随着机床的改进和诊断技术手段的提升,其知识的内容和结构是在不断完善更新的。鉴于此,本文结合人机协同技术和知识工程技术,研究面向人机协同诊断的数控机床故障诊断知识演化的核心技术,并研发滚齿机床故障知识演化及人机协同智能诊断系统,对知识演化模式、知识表示、知识推理和人机协同智能诊断技术及系统加以验证。其主要研究内容如下:针对数控机床故障人机协同智能诊断对知识演化的需求,基于对数控机床知识特点的分析,阐述数控机床故障知识演化过程和人机协同诊断所面对的一些问题,定义了面向人机协同诊断数控机床故障诊断知识演化及系统,构建了面向人机协同诊断的数控机床故障知识演化拓扑结构,阐述了该模式所包含的四核一点技术内涵、技术框架、技术特征。针对数控机床故障诊断知识表示中结构重组困难及扩展性差的问题,结合人工诊断知识特性、本体的形式化表示特点和膜计算的扩展特性,提出一种基于膜计算的数控机床本体知识表示模型。包括数控机床故障诊断本体知识表示的概念模型和关系模型,通过整合得到了数控机床故障诊断知识基础膜和知识规则约束条件,并对本体知识表示、人工知识的表示、隐性故障现象的挖掘和结构更新进行了分析。针对数控机床人机协同智能诊断系统中用于多重故障推理的知识截取困难及其验证效率低的问题,对故障知识发生的级联特性和用于推理的故障知识概念及概念关系进行了解析,通过对推理知识的简化,构建了面向人机协同诊断的数控机床多故障知识推理及其验证模型,基于该模型详细阐述了数控机床多重故障人机协同验证的过程,即通过故障诊断知识特征的整合-分离-推理验证的反复并行运算,结合故障知识逻辑推理和人工感知的高度协同,实现了人机协同诊断的数控机床多故障知识的快速准确定位,最后对数控机床多故障知识推理特性进行了分析,并通过示例进行了验证。考虑到数控机床故障信息传感器采集和人工采集在应用过程所形成的互补关系及相互验证功能和利用经验知识形成的采集方法的实用性、通用性等因素,设计了数控机床故障知识演化及人机协同诊断系统。包括构建的故障诊断知识概念、属性、关系和规则规范,并分别对面向人机协同的数控机床故障智能诊断体系架构和流程进行了设计,阐述了智能诊断的人机交互过程,并对系统评估方法进行了介绍。最后,以滚齿机床YS3120CNC6-S的知识演化与智能诊断为例,实现了滚齿机床故障知识演化及人机协同诊断系统的研发,具体包括基于本体的滚齿机床故障诊断知识收集与表示、知识库设计与实现、人机协同的滚齿机床故障智能诊断系统演示及应用。
[Abstract]:CNC machine tool is a complex system integrated with mechanical, electrical, hydraulic and air pressure. Its fault has many characteristics, such as multi source, connection, sudden and random. It can not meet the need of a large number of related characteristics and the fault diagnosis of the blind area. Therefore, the fault diagnosis of CNC machine tools is unavoidable. For the widely distributed NC machine tools, most of the fault knowledge comes from the machine tool design and manufacturing enterprises and the users. Therefore, the fault diagnosis experience knowledge has the dispersive characteristics. The fault feature analysis must rely on the related theories and other diagnostic historical experiences as a reference. Therefore, integrated man-machine collaborative knowledge, theoretical knowledge and maintenance experience knowledge are of great significance to the fault diagnosis of CNC machine tools. In particular, the fault diagnosis knowledge is improved and the diagnostic techniques are used in the process of use. The content and structure of its knowledge are constantly improved and updated. In view of this, this paper studies the core technology of the knowledge evolution of CNC machine tool fault diagnosis based on man-machine cooperative and knowledge engineering technology, and develops the knowledge evolution of the hobbing machine tool fault knowledge and the human machine cooperative intelligent diagnosis system, and the knowledge evolution. Model, knowledge representation, knowledge reasoning and man-machine Cooperative Intelligent Diagnosis Technology and system are verified. The main research contents are as follows: according to the requirement of the knowledge evolution of CNC machine tool fault diagnosis, the evolution process of CNC machine tool fault knowledge and the cooperative diagnosis of CNC machine tools are expounded based on the analysis of the knowledge characteristics of CNC machine tools. In the face of some problems, the knowledge evolution and system of CNC machine tool fault diagnosis oriented to man-machine cooperative diagnosis are defined, and the topology structure of fault knowledge evolution of CNC machine tool is constructed. The technical connotation, technical framework and technical features of this model are expounded, and the fault diagnosis knowledge of CNC machine tools is introduced. With the difficulty of structural reorganization and the poor expansibility, combining the characteristics of artificial diagnosis knowledge, the formal representation of the ontology and the expansion of the membrane computing, a model of ontology knowledge representation based on membrane calculation is proposed, including the conceptual model and relational model of the knowledge representation of the fault diagnosis ontology of CNC machine tools, and the integration of the conceptual model and the relational model of the NC machine tool fault diagnosis ontology is integrated. The knowledge base membrane and knowledge rule constraints are obtained for the fault diagnosis of CNC machine tools, and the representation of the ontology, the representation of the artificial knowledge, the mining of the hidden faults and the updating of the structure are analyzed. The knowledge interception and the efficiency of the verification are low for the multi barrier reasoning in the CNC machine tool cooperative intelligent diagnosis system. This paper analyzes the cascading characteristics of fault knowledge and the concept and concept relationship of fault knowledge for reasoning. By simplifying the reasoning knowledge, a multi fault knowledge reasoning and verification model for CNC machine tools is constructed. Based on this model, the multi fault man-machine coordination of CNC machine tools is elaborated in detail. The process of verification is to realize the rapid and accurate positioning of the multi fault knowledge of CNC machine tool, which is based on the iterative parallel operation of the integrated separation and reasoning verification of fault diagnosis knowledge, combined with the logic reasoning of fault knowledge and the high synergy of artificial perception. Finally, the multi fault knowledge reasoning characteristics of CNC machine tools are carried out at last. In view of the complementarity and mutual verification function formed in the application process of the fault information sensor of CNC machine tool and the practicability of the acquisition method formed by the experience knowledge, the failure knowledge evolution of the digital control machine tool and the human-computer cooperative diagnosis system are designed. It includes the concept of fault diagnosis knowledge, attributes, relations and rules, and designs the architecture and process of the intelligent diagnosis system of CNC machine tool fault. The process of human computer interaction is expounded. Finally, the knowledge of the system evaluation method is introduced. Finally, the knowledge of the hobbing machine tool YS3120CNC6-S is introduced. The evolution and intelligent diagnosis are taken as an example. The fault knowledge evolution of hobbing machine tools and the research and development of human machine cooperative diagnosis system are realized, including the knowledge collection and representation of the gear hobbing machine tool fault diagnosis based on the ontology, the design and Realization of the knowledge base, the demonstration and application of the intelligent diagnosis system of the hobbing machine tool fault.

【学位授予单位】：重庆大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：TG659

【相似文献】