Error Evaluation And Control Technology Of Cylindrical Grinding For Active Measurement

The performance of workpiece is affected by grinding quality.It is of great significance to improve the level of grinding automation and realize intelligent manufacturing.Active measurement technology can effectively improve machining accuracy and production efficiency by using on-line measurement and real-time adjustment.Therefore,it is very valuable to study the error evaluation and control technology for active measurement grinding.At present,there are many deficiencies in active measurement grinding in China,such as the lack of on-line evaluation of shape and position error and the lack of active measurement technology for discontinuous surface workpieces.In this thesis,the size and roundness error of workpiece are evaluated online by using active measurement system according to machining characteristics and precision requirements in different machining stages.At the same time,some data supplement methods are proposed for the discontinuous surface,and the process parameters are adjusted in time according to the evaluation results,which realizes the automatic adjustment of the grinding stage and improves the intelligence of the grinding process.The main research contents are as follows:(1)For grinding active measurement system,grinding process parameter combination is optimized in multi-objective comprehensive evaluation.A fuzzy matterelement analysis method based on composite weight is proposed to comprehensively evaluate multi-objective parameters,and the optimal grinding results are selected through correlation calculation.This method can optimize the process parameters of the platform,dynamically adjust the evaluation results according to the decisionmaking objectives,and effectively improve the grinding quality.(2)According to the characteristics of rough grinding stage and the accuracy of active measurement,the characteristic information of continuous surface is evaluated by filtering method and two-point method.An error evaluation method based on Bayesian theory is proposed to realize on-line error evaluation in rough grinding stage for discontinuous surface.(3)According to the requirements of speed and precision in fine grinding stage,the continuous surface size sequence prediction and discontinuous surface size supplement are carried out by using long and short time memory network,and the size error is adjusted according to the predicted value.At the same time,feedforward network is used to evaluate continuous surface and discontinuous surface errors efficiently.These methods are used to evaluate the size and roundness error in precision grinding.(4)After the feature information of each machining stage can be evaluated online,the size and roundness error of each machining stage can be detected in real time by using active measurement platform,and the size and roundness error fusion control can be realized by adding the setting of roundness error signal in the dimension control link.(5)A grinding active measurement and control platform is built in the laboratory.The proposed methods are verified and analyzed.The results show that the method proposed in this thesis can meet the needs of active measurement grinding and significantly improve the machining accuracy.This thesis provides a theoretical basis for the efficient evaluation of typical geometric errors in the grinding process of active measurement.It is of great significance to expand the function of active measurement technology and promote the update and application of active measurement technology.In addition,the method described in this thesis can effectively improve the intelligent degree of grinding,and also has important reference value for improving grinding accuracy and grinding efficiency.