modeling of diffusion coefficients for binary gas at p=101.325 kpa using particle swarm optimization algorithm

The diffusion coefficient of gases in a wide range of chemical processes is of great importance. semi-empirical models for diffusion coefficient prediction are useful due to their relatively lower cost compared to laboratory methods. in this study, to facilitate the equations and accelerate the calculations, appropriate models have been presented using existing parameters such as molecular mass and critical properties to determine the binary diffusion coefficient of gases. the calculations have been performed using a particle swarm optimization (pso) algorithm. this model has been used to obtain the diffusion coefficient of 84 gas dual systems at p=101.325 kpa and variable temperature (373.15-673.15 k). also, during the validation phase, the suggested model attained the most accurate prediction with r^2=0.9989. this model is capable to predict the diffusion coefficient of gases with a mean relative error percentage of 2.57% and mean square error percentage of 0.15% compared to actual data. these results are significantly better than those obtained from other models.