To investigate the laminar flow characteristics of porous media in the inner core of engine particulate filters, a two-dimensional lattice Boltzmann–Cellular Automata (LB–CA) probabilistic model is used to simulate the flow characteristics of porous media. The variation of dimensionless permeability of various numerical structures on pore scale with Reynolds number is analyzed, and the heat transfer as well as particle filtration are considered. The results show that the flow law of different structures obeys Darcy law under the condition of low Reynolds number (Re < 1). The dimensionless permeability coefficient of the ordered structure is significantly higher than that of the disordered structure; however. the filtration efficiency of the ordered structure decreases. With the increase of Reynolds number, the heat transfer increases. Furthermore, it is found that the particle size has a great influence on the filtration efficiency.