The Research Of Precision Machining Center Thermal Error's Measuring And Compensation Technology

Thermal error is one of the main factors which affects the accuracy of precision machining center. By measuring the thermal error and give out mathematical model, the affect of the thermal error can be eliminated. Based on the analysis of the sources and the form of the thermal error, this paper design a thermal error measure system for vertical precision machining center, obtained the thermal error's relatively accurate data in different working condition. By using the Grey relational analysis method, several temperature points is selected. Then use the Radial Basis Function Neural Network theory to modeling the temperature-error mathematical model, and achieved in the compensation of the thermal error. This paper researches based on these following parts:(1) Describes the background and the sources of topics, by analysis the thermal error, and combined with today's domestic and foreign thermal error compensation technology development, established research ideas and contents on this topic.(2) Analysis the physical quantity for machining center thermal error's measurement, for each physical quantity's characteristics, analysis the sensors and the measuring condition in the measurement, selected the fittest sensor. Designed fixture and temperature sensor probe structure. By analysis the sensor's output signal, select the appropriate data acquisition modules.(3) Design the spindle's thermal error measurement experiment. Debug the experimental measuring system based on the characteristics of the experimental apparatus. Then use the non-contact temperature measurement method to design the temperature points' location; Design a solution for eddy current sensor's debugging; Design different test conditions, through analysis the conditions, gives out the best thermal error measuring solution. After the experiment, gives out a preliminary analysis of the thermal error data.(4) Discussed for the reason of the machining center spindle's thermal error measuring point's optimize, choose the gray correlation analysis method to optimize the temperature points. Use three kind of data processing methods, obtained different grey sequence under different method. Then analysis the gray sequence, gives out two temperature points optimization solutions. (5) Use the RBF neural network to train the two temperature points optimization solutions, gives out the better solution. Then discussed the possibility of the program again, which combined with the actual situation, then determine the thermal error compensation temperature points.