In order to solve the problem of the low qualification rate of the pilling and coating of small-grain forage seeds, a vibration force field is introduced to the traditional vertical disk coating machine to promote the mixing of materials and improve the qualification rate of the pilling. Using the typical small-grain forage seed red clover as an example, we used the vibration force field after adding seed powder particles to a coating pot for the theoretical analysis of the force situation, using the discrete element software EDEM to construct a red clover seed simulation model with the coefficient of discretization as the evaluation index. We studied the effects of the rotational speed of the coating pot, the vibration frequency of the pot, the amplitude of the vibration of the pot, and the other operating parameters of the pot on the uniformity of the seed powder mixing, with the pelletization of the pass rate as the physical evaluation standard, using a one-way test to study the effect of operating parameters on the pelletization pass rate. We used the qualified rate as the physical evaluation standard, through a single-factor test, to study the influence of the working parameters on the qualification rate of the pilling. The results show that the simulation test evaluation index of the discrete coefficient and the physical test evaluation index of the pilling qualification rate with the change rule of the working parameters is consistent with the discrete coefficient, and can be used as an indirect evaluation index of the quality of pilling. To further determine the optimal combination of working parameters, a quadratic regression orthogonal design test was conducted with the discrete coefficients as the evaluation index, and the second-order regression equations of the red clover seeds were established and analyzed by ANOVA using Design-Expert software. The study shows that, when the rotational speed of the coating pot is 307.204 rpm, the vibration frequency is 2.526 Hz, and the vibration amplitude of the coating pot is 5.843 mm, the predicted coefficient of dispersion at this time is 8.1%. Simulation using the best combination of parameters to obtain the average value of the dispersion coefficient of 8.4%, relative to the predicted value of 3.7%, indicated that the optimization of the experimental regression model is accurate, and the results obtained for the vibration of small seeds under the conditions of the design of the pellet granulation coating machine and the optimization of the pelletization coating process parameters have a certain degree of reference significance.