Abstract: In order to achieve the high efficiency and precision of milling process, it is very important to study the reasonable analysis and control method of machining deformation. Consequently, the finite element model is reasonably created to simulate the milling process of thin-walled workpiece. The milling of T thin-walled workpiece can validate the agreement of simulated values with the corresponding experimental data. And then, the finite element method is used to obtain the training samples in addition that the orthogonal experimental design method is adopted for the input samples. The genetic algorithm can skillfully be developed to formulize the back propagation (i.e., BP) neutral network model by taking the prediction error of output samples as the individual fitness. It can predict the machining deformations of workpiece with less than 6% relative error. Finally, an optimal model of tool parameters is further suggested for the minimum machining deformation. In light of the principle in which the smaller the machining is deformation the stronger the chromosome is, the genetic algorithm can be employed to solve the optimal model of tool parameters. The presented method can realize the optimal design and selection of milling tool based on the improvement of the calculation efficiency.