Exercise load is a quantitative physiological index and training intensity index that athletes and coaches are extremely concerned about. In order to obtain valuable potential information from a large amount of random exercise load data, with the assistance of data mining technology, a training load monitoring method based on aerobics athletes is proposed, which utilizes the association rule method to mine valuable information for predicting the amount of human exercise. After that, a kinetic model of the sports training function monitoring system with mutual feedback relationship among athlete’s function, training and recovery is constructed to provide rationalized suggestions for training load regulation of aerobics athletes. Numerical simulation results show that the system in this paper is more suitable for data mining of athletic biochemical indexes of aerobics athletes. The numerical simulation results of aerobics training load regulation show that with the gradual increase of the training load level and training movement intensity regulation factor of aerobics athletes, their recovery cycle optimization factor will gradually decrease, and when each regulation factor is increased by 20%, the athlete’s athletic ability level can be increased by up to 7.87% compared with the original. In addition, the regulation of movement complexity shows that the training load has a time lag effect with the improvement of athletic ability level.
