PANews reported on November 4 that the Center for Complex Decision Analysis at Fudan University released a prediction of the 2024 US presidential election based on computer simulation. The results showed that Trump is very likely to win the US presidency again (with a probability of more than 60%). The research team said that the prediction is mainly based on the simulation technology of behavioral agent modeling (ABM). Our prediction is divided into three steps: (1) Through simulation prediction based on ABM, we obtained the distribution of votes and the ownership of electoral votes in 8 key states in the United States (due to budget constraints, we selected 8 key states); (2) Through 1 million rounds of Monte Carlo simulation, we obtained the tracking poll prediction results and the ownership of electoral votes in the remaining 42 states and Washington, DC; (3) The two results are summarized to finally obtain the overall prediction results of the 2024 US election. Our ABM prediction covers eight key states (these eight states are the states where the gap between the Democratic Party and the Republican Party is less than 5% in the 2020 election). The specific prediction results of the eight key states (based on ABM simulation prediction) are as follows. Trump will win: Michigan, Ohio, Pennsylvania, Florida, Wisconsin. Harris will win: Georgia, Arizona, North Carolina. Given the special attention this year's election has attracted, the Center for Complex Decision Analysis at Fudan University said it will release technical details after the election results.