According to the analysis of mechanism of sintering process, the prediction of permeability is verified to having the character with uncertainty in the prediction model and the input parameters. The Takagi-Sugeno method combined with mechanism analysis is employed to resolving the uncertainty. In order to avoid a great deal of computation and non-identifiable problem possibly presented in the traditional identification method, the hybrid particle swarm algorithm with gradient acceleration is applied to identify the parameters of Takagi-Sugeno. The simulation results show that the proposed method can overcome the uncertainty successfully, predict the permeability accurately, and have the robust character to a certain degree.