Since the diffusion coefficient is a key parameter to characterize the diffusion rate of methane molecules, its measurement and solution have always been a research hotspot. The diffusion coefficient is normally solved through analytical solutions of theoretical models, which is complex and poorly applicable. In comparison, the numerical simulation optimization method can seek a solution easily and quickly, providing a clue for solving such problem. In this paper, first, gas desorption experiments were conducted on coal samples with different initial gas equilibrium pressures, coal particle sizes, and metamorphic degrees. Combined with existing theoretical models, the numerical simulation optimization method was adopted to solve the diffusion coefficient of the coal particle. Furthermore, the applicability and advantages of the numerical simulation optimization method were discussed. Finally, the variation law of the diffusion coefficients was analyzed. The results demonstrate that the numerical simulation optimization method can not only solve the diffusion coefficient easily and quickly but also reveal the law of diffusion concentration with time. The d values between the solution results and the experimental data under different conditions are all smaller than 0.2, which proves the effectiveness and accuracy of the simulation optimization method. The diffusion coefficient of gas from coal particles is unrelated to the initial gas equilibrium pressure, yet it has a Z-shaped relationship with the coal particle size and a V-shaped relationship with the metamorphic degree.