In this paper,the plastic front module of a certain vehicle is studied.Based on the elastic modulus E,the Poissons rati-o U and the density P as optimization variables,the first three order modal test results are used to optimization objectives,the multi-is-land genetic algorithm is the optimization strategy for multi-variable and multi-objective numerical model.The radial basis functions neu-ral network is used to construct the numerical model to reduce computational time,and the feasibility is verified by the R-Squared reli-ability analysis,the root mean square of the modal frequency are 0.9993,0.9993 and 0.9992 respectively.The results show that the corresponding simulation modal frequencies with the experience values of the general material parameters are 17.23Hz,33.10Hz and 41.37Hz,the deviations with 17.9072Hz,35.8126Hz and 44.4983Hz of test results are 3.78%,7.57%and 7.03%respectively; the corresponding simulation modal frequencies with the inverse material parameters are 18.44Hz,35.45Hz,and 44.58Hz.The devia-tions are 2.98%,1.01%and 0.18%respectively.