The ANSYS Workbench module is used to analyze the modal of the machine tool stand column based on a parametric modeling of the column conducted by UG.The first modal frequency was found to be relatively low.The dimensional parameters which have greater impact on the first modal frequency and mass of the column were found out by using sensitivity analysis on dimensional parameters of column. Accordingly, a basement about the changing dimension and the corresponding column properties was established. On this account, a neural network model was built to obtain corresponding relationship between the dimensional parameters of column with the first modal frequency and mass.The genetic algorithms were worked with different optimization target.Finally, an appropriate optimization scheme emerges by comparing the results.