Abstract: It is very important to develop breeding techniques for breed improvement and grain production increase. However, the most breeding experiment is completed by labor, and it has the disadvantages of high labor intensity, low efficiency and high cost, and because of the human factors and personal differences, the quantity and accuracy of plot breeding experiment is seriously affected, and hampered the smooth proceed of breeding. So the breeding seeding machine plays an important role in improve work efficiency and breeding quantity. There are many varieties of soybean in plot breeding experiment and they are always different in size and shape, as such, the variety adaptability for the plot breeding seed metering device become a key factor. So the mathematical model of type hole structure of socket-roller seed plate for soybean should be created. In our study, firstly, six species of soybean in Heilongjiang were selected as experiment materials and a statistical analysis of their physical parameters was conducted. Based on this, the ratio of diameter, depth and chamfer length of the type hole to the average diameter of soybean seed was determined by the theoretical analysis of seed’s filling process. The discrete element simulation model of socket-roller seed plate was established and simulated the process of seed metering by EDEM. In order to get the optimal size parameter of type hole, the parameter optimization of type hole simulation experiment was conducted, and the ratio of diameter (DI Ratio), depth (DE Ratio) and chamfer length (CH Ratio) of the type hole to the average diameter of soybean seed was selected as experiment factors. The single-seed rate, multi seed rate and empty-seed rate was selected as experiment index, quadratic regression rotational combination experiment was done and the data was analyzed by Design-Expert 8.0.6 software, and we then obtained the regression model and analyzed the influence of factors on the index, the experiment results indicated that for single-seed rate, the influence of DI Ratio was very significant (P<0.01), the DE Ratio and CH ratio was significant (P<0.05), the single-seed rate first increased and then decreased with the increase of the DI Ratio, DE Ratio and CH Ratio, for multi-seeding rates, the influence of the DI Ratio and DE Ratio was very significant (P<0.01), and CH Ratio had no effect on multi-seed rate (P>0.05). The multi-seed rates increased with the increase of DI Ratio and DE Ratio. For the empty-seed rate, the influence of the DI Ratio and DE Ratio was very significant (P<0.01), and CH Ratio had no effect on empty-seed rate (P>0.05). The empty-seed rate decreased with the increase of DI Ratio and DE Ratio, the empty-seed rate and multi-seed rate were all decreased first and then increased with the increase of CH Ratio. The single-seed rate was more than 90%, multi-seed rate and empty-seed rate was less than 5% by the parameter optimization module of Design-Expert 8.0.6. When the CH Ratio was 0.14, the range of DI Ratio was 1.73 - 1.91, and the range of DE ratio was 0.76 - 1.25, the seed plate had the best performance. In accordance with the optimal parameters, varieties of Ken bean 40 and Ken Feng 17 were selected as experiment materials, planetary gear type transposition type hole seed plates were made by 3D Printing with the optimal type hole, and meanwhile the other two seed plates with different type hole without the optimal parameter were selected as contrast. The comparative experiment was done and the results showed that planetary gear type transposition type hole seed plate with the optimal type hole parameter had the best performance. The results were basically consistent with the simulation results, and the simulation optimization model had been verified. Meanwhile, the research provide a reference and method for the design and optimization of the type hole of seed plate.