In order to simulate the morphology and structure of mono-crystalline silicon under different etching conditions such as temperature, concentration, with or without surfactant, the atomic structure model of mono-crystalline silicon is constructed and the relationship between the etching rates of main crystal planes and the corresponding atomic structure are analyzed. A surface atomic etching function (Si-RPF) suitable for mono-crystalline silicon etching simulation is proposed. The numerical relationship between the macroscopic etching rate of crystal planes and the microscopic removal probabilities of atoms is clarified and a Kinetic Monte Carlo (Si-KMC) anisotropic wet etching process model is constructed based on genetic algorithm. Based on the step flow theory, the process model can explain the cause of the anisotropy of silicon etching from the perspective of microscopic atoms, clarify the role of different types of atoms in the etching process, and realize the accurate simulation of the three-dimensional etching morphology of silicon substrate under different etching conditions. By comparing the experimental data and simulation results of silicon etching with or without the addition of surfactant, the simulation results of Si-KMC etching process simulation model can reach more than 90% simulation accuracy.