The grouping and optimization approach to identify the key thermal points on machine tools is studied.To solve the difficulty in grouping because of the high correlated variables from distinct groups,the variables gro...The grouping and optimization approach to identify the key thermal points on machine tools is studied.To solve the difficulty in grouping because of the high correlated variables from distinct groups,the variables grouping technique is improved.Temperature variables are sorted according to their relativities with the thermal errors.The representative temperature variables are determined by analyzing the variable correlation in sort order and removing the other variables in the same group.Considering the diverse effect of importing the different variables on thermal error model,the method of variable combination optimization is improved.Regression models made up of different combination of representative temperature variables are evaluated by the index of both the determined coefficient and the average residual squares to select the combination of the temperature variables.For the machine tools with complicated structures which need more initial temperature measuring points the improvement is demanded.The improved approach is applied to a precision horizontal machining center to identify the key thermal points.Experimental results show that the proposed approach is capable of avoiding the high correlation among the different groups' variables,effectively reducing the number of the key thermal points without depressing the prediction accuracy of the thermal error model for machine tools.展开更多
基金Sponsored by the Special Fund for Scientific and Technological Achievement Transformation of Jiangsu Provincethe Basic Scientific Research Professional Expense of NUAA for Special Project
文摘The grouping and optimization approach to identify the key thermal points on machine tools is studied.To solve the difficulty in grouping because of the high correlated variables from distinct groups,the variables grouping technique is improved.Temperature variables are sorted according to their relativities with the thermal errors.The representative temperature variables are determined by analyzing the variable correlation in sort order and removing the other variables in the same group.Considering the diverse effect of importing the different variables on thermal error model,the method of variable combination optimization is improved.Regression models made up of different combination of representative temperature variables are evaluated by the index of both the determined coefficient and the average residual squares to select the combination of the temperature variables.For the machine tools with complicated structures which need more initial temperature measuring points the improvement is demanded.The improved approach is applied to a precision horizontal machining center to identify the key thermal points.Experimental results show that the proposed approach is capable of avoiding the high correlation among the different groups' variables,effectively reducing the number of the key thermal points without depressing the prediction accuracy of the thermal error model for machine tools.
