Research On On-Line Inspection And Error Compensation Technology Of CNC Machine Tools

With the continuous development of manufacturing technology and the continuous advancement of technology,people have higher and higher requirements for the quality and precision of some complex processed parts and the need to improve the inspection efficiency of products.The inspection technology based on coordinate measuring machines has been quite mature.However,there are still some problems on the inspection direction,such as the positioning errors caused by relocation which are difficult to eliminate,larger parts can not be inspected.A closed-loop machining inspection system combines machining and inspection on a machine tool,which is called on-line inspection of CNC machine tools.on-line inspection is also a hot issue in domestic and foreign research fields in recent years.The inspection probe is mounted on the main shaft,and the signal receiver transmits the data,which not only avoids the positioning error of the secondary clamping but also reduces the time required for the disassembly and handling of the parts.On-line inspection of CNC machine tools improve the inspection accuracy and efficiency of the parts,and this method achieves an integrated mode of processing,inspection and compensation.Due to the complexity of the inspection data processing and the length of the inspection process,the Gaussian process regression method in supervised learning can use the input and output in the sample to get the relationship between them.According to the corresponding relationship between input and output,the unknown result can be obtained.Based on some problems in the on-line inspection of complex curved parts,this paper deeply studies key technologies such as interference avoidance simulation,numerical control system on-line inspection error compensation and Gaussian process regression intelligent algorithm,and improve the inspection accuracy of complex curved parts by simulation and error compensation techniques.The intelligent algorithm of Gaussian process regression solve the inspection efficiency and the complexity of data processing.The main research contents of this paper are as follows:(1)According to the sophisticated inspection technology of complex curved parts,UG secondary development technology and Gaussian process regression algorithm,they are conducted in-depth research on its research status.On-line inspection technology,off-line inspection technology,error compensation technology,UG secondary development of interference avoidance technology and Gaussian process regression have been comprehensively analyzed,then they are found their advantages and problems.(2)In order to prevent the interference between the probe and the curved surface during the actual operation,firstly,the interference module of UG is analyzed,and the inadequacies of the module are pointed out,and the developed arc length measurement point generation system is used to generate a series of test data points on the surface.The working principle of the simulation probe and the causes of the interference and how to implement the interference phenomenon are introduced in detail.On the basis of the interference problem,how to realize the automatic obstacle avoidance function of the probe is extended.Finally software is verified by simulation experiments.(3)In order to improve the on-line inspection accuracy of CNC machine tools,various factors affecting the error accuracy of the contact inspection system and how to compensate the error are analyzed.The main error sources of the inspection system,such as the positioning error of the machine tool,the pre-travel error of the probe and the error of the probe radius,are analyzed in detail.Based on the comprehensive modeling of the three-axis CNC machine tool,the laser interferometer is used.Based on the measurement and compensation of the positioning error of the three-axis CNC machine tool,a compensation method based on cubic spline curve is proposed.Based on this research,a CNC machine tool positioning error compensation system is developed by VC++6.0 software development platform.The CNC machining,measurement and compensation experiments of the test and complex curved parts are verified for accuracy,and the On-line inspection data are compared with the results of the CMM test data.(4)According to the problem of complex surface part inspection efficiency,this paper proposes an artificial intelligence algorithm based on Gaussian over regression.The detailed regression method of the Gaussian process is discussed,and how to optimize the Gaussian process parameters and the kernel function are discussed.A classic example shows how to perform Gaussian process prediction.Finally,by constructing a Gaussian process regression model,the theoretical points of the complex surface are utilized.And the limited data points of the measurement data are used for prediction,and the prediction results are compared with the actual measurement results,which better predicts the unknown data and greatly improves the inspection efficiency of the surface.Finally,this paper summarizes the full text and points out some problems that need to be solved in the future.