Research On Wheel Wear Condition Monitoring System Of Bearing Ring In Cylindrical Grinding

High end bearing is the core basic part for safe and stable operation of all kinds of highend equipment.As one of the main processing methods of bearing ring,cylindrical grinding has a complex processing mechanism.The wear state of grinding wheel is directly related to the quality of grinding workpiece.In dynamic grinding,the coupling and complex relationship between various influencing factors make it difficult to build the dynamic model of grinding wheel wear directly from the mechanism.Therefore,from the perspective of indirect monitoring,this paper deeply studies the recognition technology of multi-sensor information fusion of grinding wheel wear state,and develops the monitoring system of grinding wheel wear state in the grinding process,so as to provide theoretical basis and technical support for the digital,networked and intelligent processing of high-end bearings in China.The specific research work includes the following contents:Firstly,carry out the grinding wheel wear experiment of bearing ring cylindrical grinding,and repeat the experiment with constant grinding parameters until the grinding wheel is seriously worn.By building a multi-sensor information acquisition platform in the grinding process,the sensor information such as AE(acoustic emission),vibration and power can be obtained in real time,and the surface roughness of the machined workpiece and the surface morphology of the grinding wheel can be measured in situ.The sensor information of grinding process is combined with the result information of grinding wheel detection to comprehensively describe the grinding wheel wear process,so as to provide practical and reliable experimental data for the subsequent research of grinding wheel wear condition identification.Secondly,according to the experimental data of grinding wheel wear,multi-sensor signal analysis and grinding wheel wear feature extraction are carried out.Around the multi-sensor information processing flow,the signal preprocessing removes the interference of external factors and makes the signal energy more concentrated;Signal feature extraction converts complex digital signals into relatively simple eigenvalue description;The feature optimization of signal selects high correlation features,and the feature dimension reduction reduces the redundancy of the optimal feature set,so as to obtain a simplified subset of grinding wheel wear features,which makes it possible to build a stable and reliable grinding wheel wear condition recognition model.Thirdly,combining artificial neural network with statistical theory,a grinding wheel wear condition recognition model based on multi-sensor information hierarchical fusion is constructed to deeply mine the grinding wheel wear information.BP neural network is used to establish the feature level recognition model of grinding wheel wear condition.Through model training and testing,the recognition accuracy of the model is 83.3%,and the support rate of grinding wheel wear recognition proposition is close to 70%.In order to improve the recognition accuracy and recognition accuracy,the D-S evidence theory is used to establish the decision-making level recognition model of grinding wheel wear condition.The recognition results of the model show that the D-S decision-making model integrates the uncertain components of each BP sub network model,eliminates the influence of uncertain components on the recognition result,improves the recognition accuracy,and the reliability support of each recognition proposition exceeds 93%,which has high recognition accuracy of grinding wheel wear condition.Finally,based on the theoretical research foundation,combined with Java and other advanced programming languages,the grinding wheel wear state monitoring system is developed to realize the functions of grinding process state information acquisition,machining state identification,data management and service.Through the relevant technical analysis,demand analysis and architecture design of the monitoring system,and using MYSQL,Spring Boot,Layui and other framework technologies,the back-end management and front-end page design of processing data are realized to ensure the normal interaction of front and rear data.The system not only realizes the basic information query and physical data management of grinding,but also can monitor the wear state information of grinding wheel,which can meet the needs of remote service of grinding information.