At present, the simulation results of athletes' running training intensity are only a single screen, which leads to the low visibility and long simulation time of athletes' running training intensity, which are inconsistent with the actual results of athletes' running. For this reason, it is proposed to use the images reflected between frames to define the athletes' running training intensity. simulation. Use statistical measurement methods to obtain athletes’ physiological parameters, and calculate matching factors to match athletes’ physiological parameters with those of the simulated human; determine the simulated human’s joint rotation, rotation angle, running step length, pace, pace and speed on the simulation platform Athletes are consistent, calibrate the motion posture of the simulated person; substitute the calibrated motion posture data into the simulation platform, and use the images reflected between frames to form a continuous virtual animation of athlete running training, analyze the athlete's running training, and define the athlete's running training intensity. The experimental results show that under the same experimental parameters, the consistency between the simulation results and the real running motion model of the human body is detected. It is concluded that the proposed method can accurately simulate the athlete's running posture, with high visibility and short simulation time, which is consistent with the actual running movement of the athlete. The results are consistent, the simulation effect is good, and it can provide some help for supervising the athletes' running training process.