Simulation has become a powerful method for military research and combat training due to its intuitive visualization, repeatability, and security in contrast to real-world training. Previous studies often divided cognitive and physical factors into isolated models using separated platforms. Ideally, both cognitive and physical aspects of a virtual soldier should be modeled on the same platform. We demonstrated an integrated modeling that combines cognitive models with physical human models. A simple task was used, requiring the virtual soldier to navigate in a virtual city, avoid enemies, and reach the destination asap. The Queueing Network-Adaptive Control of Thought Rational cognitive model helps the virtual soldier make choices after encountering enemies. Based on the information collected, the soldier will choose different strategies. Two general-purpose methods from the cognitive modeling and digital human modeling were combined. The results were able to capture the behavioral states as planned and visualize the movement of the virtual soldier, who was able to complete the task as expected. The results demonstrated the feasibility of integrated models combining cognitive and physical aspects of human performance in the application of virtual soldiers. Future studies could further compare the results of model output with human empirical data to validate the modeling capabilities.