Research On Knowledge Modelling Methods And Configurable Knowledge-based System For Fault Diagnosis Of Machine Tools

Fault diagnosis is a critical activity in health management of machine tools due to its great significance in such efforts as prolonging lifespan,improving production efficiency,and reducing production costs.From the current literature study,there are many studies on data-driven method such as signal processing,fault recognition,physical model and so on.While knowledge-driven methods,such as knowledge representation,management,application and sharing,have attracted less research attention.Knowledgedriven methods are effective in improving the knowledge representation ability and intelligence level.As the complexisty of machine tools,demand for intelligence level and development of artificial intelligence,there is an urgent requirement for intelligent fault diagnosis system.Therefore,this paper investigates on the knowledge modelling methods and configurable knowledge-based systems for fault diagnosis of machine tools.The main contents are as following.Firstly,a multi-perspective analysis model is proposed for machine tool’s fault diagnosis domain.It presents a comprehensive view of fault diagnosis in machine tools from three dimensions,namely hierarchy,activity and time.Then,a knowledge management architecture of fault diagnosis for machine tools is presented.It takes knowledge as a core to acquire,represent,organize and utilize the static and dynamic knowledge produced in each hierarchy,life stage and fault diagnosis activity,enables implementing knowledge-driven,data-driven and hybrid diagnosis and thus achieves the sharing,integration,application and update of various fault diagnosis knowledge.Secondly,a knowledge modelling methodology for fault diagnosis of machine tools based on formal semantics(KMM-MTFD)is proposed in this paper to build an open,shared,and scalable ontology-based knowledge model of fault diagnosis of various machine tools(OKM-MTFD).A predicate-logic-based analysis method of fault elements is proposed to study the fault diagnosis domain and extract the common domain knowledge,which enables the establishment of the core ontology of OKM-MTFD to assure formal semantics.Then,using the proposed two-stage classification method of fault elements and external ontology reference methods,the core ontology can be extended into OKM-MTFD for a type or a specific machine tool.Knowledge reasoning and querying methods are provided to utilize the knowledge base efficiently.An ontology-based knowledge model and knowledge base of a hobbing machine tool is presented to exemplify the validity of the proposed KMM-MTFD.Thirdly,a hybrid fault diagnosis method for mechanical components is proposed based on ontology and signal analysis(HOS-MCFD).It is a systematic approach covering the whole process of fault diagnosis: feature extraction from raw data,fault phenomenon identification and fault knowledge modeling and reasoning using ontology and semantic web technology.A semantic mapping approach is presented to relate signal analysis results to ontology elements.It integrates the advantages of signal analysis and ontology.It can be applied to deal with fault diagnosis more accurately,systematically and intelligently.This method is assessed with vibration data of rolling bearings.The experimental results prove the proposed method effective.Then,a configurable method for fault diagnosis knowledge of machine tools(CMFDK-MT)is proposed in this paper.For machine tool manufacturers,a general fault diagnosis method and a software framework are needed to construct fault diagnosis systems for various machine tools and fault types,or the same type of machine tools under various life cycles,working conditions and operating environments.Firstly,an ontologybased fault diagnosis method for machine tools and an improved fault diagnosis process are introduced.Then,a configurable fault diagnosis platform for machine tools(CFDPMT)is designed.CFDP-MT supports explicit knowledge representation with formal semantics,efficient knowledge utilization,efficient integration of various fault diagnosis methods and technologies and configuration of fault diagnosis activies.Then,the configuration approaches for fault diagnosis activities,namely fault detection,identification,diagnosis and solving,are studied respectively.The configuration and implementation methods of the CFDP-MT framework are also presented.A configurable method for data acquisition of machine tools(CMDAQ-MT)is proposed.It supports for various data sources and data acquisition requirements and provides uniform operating data of machine tools.It provides a uniform framework including a common manmachine interface,data communication protocol and plugin interface.The obtained raw data will support accurate fault diagnosis.Finally,a prototype fault diagnosis system is constructed for a CNC hobbing machine tool and its manufacturer.Experiments,such as fault code,rolling bearing and gear,are carried out to verify the effectiveness of the proposed methods.