Solidification microstructure of aluminum alloy has a great influence on the properties of the casting. An aluminum alloy with the structure of fine equiaxed grains has low casting defects and presents excellent mechanical properties. Recently, chemical inoculation by adding grain refiner is the technique most extensively used to achieve a fine, equiaxed grain structure of Al alloy in the industrial production. In order to investigate the detailed solidification microstructure evolution of the alloy, many numerical models have been proposed. Cellular automaton method is one of the powerful tools for simulating the morphology evolution of detailed grains in the solidification process of alloy. However, the present cellular automata model has a shortcoming, that is, its calculation of the nucleation rate is based on the experimental number density of grains. In this work, a population dynamics-cellular automaton model is developed for describing the solidification microstructure evolution of the inoculated aluminum alloy. The model takes account of the heterogeneous nucleation of α-Al nucleus, the initial spherical growth of <i>α</i>-Al grains and the dendritic growth process. The model is used for simulating the solidification microstructure evolution of the commercial-purity aluminum (CP-Al) inoculated by Al-5Ti-1B master alloy. The results indicate that the heterogeneous nucleation process of α-Al can be divided into the two stages. In the early stage of nucleation, there are enough effective particles in the melt. The nucleation rate of <i>α</i>-Al increases with the increase of the undercooling of the melt. After a short time, the nucleation of α-Al is dominated by the number density of the effective particles in the melt. Nucleation process stops when the recalescence takes place. The effects of the additive amount of Al-5Ti-1B master alloy and the cooling rate of the melt on the solidification microstructure of the CP-Al are investigated by using the established model. The final solidification structures of CP-Al are predicted. And a comparison between the predicted results and the experimental ones shows that they are in good agreement with each other.