This paper constructs a model for an athlete's body heat recovery and reuse system. A flexible distributed architecture is designed, consisting of a body heat capture layer, an adaptive transmission network, a phase change energy storage core, and a multi-energy conversion terminal. Key components include carbon nanotube/organic silicon composite membranes, polyimide microtubes, and n-heptadecane/graphene PCM. A simulation model was developed using ANSYS FLUENT to simulate system performance at varying exercise intensities and compare the results with experimental results. The results show that the experimental heat storage efficiency under moderate-intensity conditions is 82.3%, while the simulated value is 84.1%, with an error of 1.8%. The heat loss curves for the transmission pipeline agree well with the simulation curves by 96%. During high intensity exercise, the experimental entropy production rate is 0.0085 J/Ks, while the simulated value is 0.0089 J/Ks, with an error of 4.7%. The efficiency in heating mode is approximately 89%-91.5% at 25°C ambient temperature across moderate-to-high exercise intensities (500-700W input), while in power generation mode, it is approximately 4.7%-5.1% under 6-10°C temperature differences between the storage unit and environment. The experimental and simulation errors were both within 8%, validating the model's applicability and providing a theoretical and experimental basis for the efficient recovery and utilization of body heat.