In the process of CO2 injection development in heavy oil reservoirs, asphaltene exhibits complex behaviors such as precipitation, flocculation, and deposition, significantly affecting the flow channels within the reservoir and influencing the entire lifecycle of reservoir development. However, most current numerical simulation studies fail to adequately consider the impact of asphaltene deposition, leading to inaccurate production forecasts and difficulties in quantifying and optimizing injection and production parameters scientifically. This study constructs a numerical simulation research framework with experimental data based on the Petrel platform to simulate the CO2 injection development process in the Bei 10 oilfield, Junggar Basin (China). Experimental results indicate that the crude oil in the oilfield has a high asphaltene content and its sensitivity to CO2 injection is within the normal range. An asphaltene precipitation model was established based on experimental results and applied to the CO2 injection simulation in this region. The simulation results shows that cyclic CO2 injection development is more effective than continuous CO2 flooding, with the optimal injection rate at 50 t per day, optimal injection volume at 500 t, optimal soaking time at 20 days, and recommending 3–4 injection cycles.