Research On Key Technologies Of Health Warning And Fault Diagnosis For Five-axis Tool Grinding Centers

As a high-end machine,the five-axis tool grinding center has been widely used in tool production and other industries.However,compared with foreign equipment,it has low reliability and high maintenance cost.Therefore,research on data acquisition,condition monitoring,health warning and fault diagnosis will help to improve the equipment utilization and intelligence of the five-axis tool grinding center,reduce the equipment maintenance loss,and enhance the international competitiveness of the domestic five-axis tool grinding center.Study the data acquisition technology of the five-axis tool grinding center.According to the historical fault data and the structural characteristics of the equipment,the common fault phenomena and fault types are summarized,and the main fault causes are analyzed to determine the key monitoring parameters.Based on the data transmission protocol of CNC system,and the external vibration sensor,the data acquisition scheme of equipment operation status is established,the data acquisition program is programmed,and the verification experiment is carried out.The processing status data of the equipment is successfully collected,and the feasibility of the scheme is verified.Establish the five-axis tool grinding center condition monitoring system based on B/S architecture.The webpage interface is based on the design status of the react,and functions such as user login verification,authority management,machine information management,and real-time display of processing status are realized.Based on database requirements analysis,using TimescaleDB optimized for time series data to build a database for data storage and query testing,the result proved that the average query time is 31.3% shorter than PostgreSQL.A five-axis tool grinding center health warning and fault diagnosis plan is proposed.Aiming at the high warning value and lack of pertinence of the existing early warning mechanism,based on the collected state monitoring data,a health warning scheme based on the spindle temperature of the equipment is proposed.The spindle bearing health warning model based on support vector machine is established by collecting the vibration signal of the spindle bearing.Further,a fault diagnosis scheme based on multi-source information fusion technology is established.According to the data characteristics of each axis speed,follow-up error,temperature and load rate,the feature quantity extraction technology is studied to regularize and reduce the feature quantity.Based on the BP neural network a fault diagnosis model is testablished.According to the verification test of the diagnostic model,the fault identification correct rate reached 97%.