Research On Fault Knowledge Evolution Technologies Of Human-machine Synergy Diagnosis For CNC Machine Tools

The CNC machine tool is a complex system which integrates mechanical, electrical, hydraulic, pneumatic, etc. Its fault diagnosis has many characteristics, such as multi source, correlation, sudden and random. Due to the characteristics of the correlation property of parts and the monitoring blind sports, it cannot meet the needs of the CNC machine tool’s fault diagnosis only by the sensing device. Therefore, the method of human-computer cooperation is the inevitable choice for the fault diagnosis. For the CNC machine tools with widespread distribution, its most of the fault knowledge is derived from the design & manufacturing enterprises and users of the machine tools, so that the fault diagnosis experience knowledge has the characteristics of dispersion. In the process of fault diagnosis for CNC machine tools, the fault feature analysis must rely on the relevant theory and other diagnostic historical experience. Therefore, it is significant to integrate the knowledge of human knowledge, theory knowledge and maintenance experience in the fault diagnosis of CNC machine tools. Especially, the integration of knowledge in the use process, with the upgrading of machine tools and diagnostic techniques, the content and structure of its knowledge is constantly improving and updating.In view of this, this paper combines with human-machine collaborative technologies and knowledge engineering technologies, and studies on the core technologies of fault knowledge evolution technologies of human-machine synergy diagnosis for CNC machine tools. And the knowledge evolution and intelligent diagnosis system for the hobbing machine tools is developed for verifying the technologies and theories of the knowledge evolution model, the knowledge representation, knowledge inference and the human-computer synergy intelligent diagnosis system. The main researches of this paper are follows:Focus on the demand of the knowledge evolution for the intelligent fault diagnosis of the CNC machine tools by human-computer synergy, this paper analyzes the knowledge characteristics of numerical CNC machine tools, and reveals some problems in the process of fault knowledge evolution and human-computer synergy diagnosis of the CNC machine tools. The knowledge evolution and system of fault diagnosis for CNC machine tool for human machine collaborative diagnosis is defined. And the topological structure of the fault knowledge evolution technologies of human-machine synergy diagnosis for CNC machine tools is proposed. This paper describes the connotation of technology(named: theory of four Core and one Point), technical framework, technical characteristics, knowledge inference rules and system design requirements.Aiming at the difficult problems of the knowledge restructured and poor scalability of the knowledge representation, this paper presents the method of ontology knowledge representation based on membrane computing of CNC machine tool’s fault diagnosis, which combines characteristics of the knowledge of artificial diagnosis, the formal description of ontology and the expansion of membrane computing. The basic membrane and the knowledge rule constraint are obtained by integrated the concept model and relationship model for the CNC machine tool’s fault diagnosis. And the knowledge representation of ontology, the artificial knowledge representation, the mining of the hidden failure phenomenon and the structural update are analyzed.This paper aims to solve the difficulties in knowledge simplification and inefficiency in information verification, which improves the accuracy and effectiveness of multi-fault knowledge reasoning in expert systems, in accordance with the Production Rules and Dempster-Shafer(DS) theory in knowledge presentation and inference of rule-based systems. It describes concepts and their relations of the fault knowledge and simplifies the fault knowledge inference model, where the multi-fault knowledge inference and verification model is proposed. This paper elaborates machine tools multi-fault knowledge reasoning and verification process through fault diagnosis knowledge integration-separation-inference and information verification of repeated parallel computing to achieve multi-fault diagnosis by human-computer interaction. It realizes the multi-fault diagnosis fast and accurate positioning. At last, the analyses of the model are explained.Taking into account the complementary relationship and mutual verification function of CNC machine tool fault information collection and the method of mutual verification and mutual verification, this paper designs the fault knowledge evolution and human-computer cooperative diagnosis system of numerical control machine tools. The concepts, attributes, relations and rules of fault diagnosis knowledge are described, and the architecture and process of intelligent fault diagnosis system for CNC machine tools are designed. This paper introduces the human-computer interaction and the system evaluation.Finally, this paper takes the technologies of intelligent fault diagnosis and knowledge evolution of the gear hobbing machine tool(YS3120CNC6-S) as the research object, and introduces fault knowledge evolution process and the system of intelligent diagnosis design and application, including the knowledge collecting and representation based on ontology, knowledge base designing and implementation, demonstration and application of intelligent fault diagnosis system for gear hobbing machine tools.