Abstract: The development task of high-end equipment has many characteristics, for example, the number of influencing factors is huge, the uncertainty is great and the network structure is complex. In order to deal with this problem, a Monte Carlo simulation method is used to simulate the execution process of the task network. It can estimate the duration, cost, rework times, and failure rate. Considering the uncertainty of the duration and the cost of subtasks, the relationship between the subtasks and the self-loop, rework iteration, overlap, failure and learning effect in the implementation process, our method improves the rules of resource conflicts. Based on the design structure matrix, the simulation model is established. After that a simulation algorithm is implemented. The data of the development task of an uninhabited combat aerial vehicle is studied. Then this paper analyzes the data such as the duration, cost, rework times, and failure rate of the simulation output. Finally, the data is compared with the data in the related literature, which shows that our method works well in this case.