In this research paper, the parameters of body ability training, that is, the quantity of exercise, the strength of exercise, the length of exercise, and the times of exercise, are defined by us as exercise load parameters. After that, we make use of the genetic algorithm to carry out the optimal distribution and adjustment of these parameters. An improved hybrid genetic algorithm (SGA) with global optimization capability was designed by combining the genetic algorithm with the simplex method, and the optimal allocation of exercise load parameters was optimally designed using this algorithm. Then NSGA-II was used as the basic search strategy to design a regulation model of exercise load parameters based on the feedback regulation mechanism. Finally, the two are combined to construct a physical fitness training prescription generation system that satisfies user preferences. It is verified that SGA and NSGA-II have better convergence and optimization performance than traditional genetic algorithms.The experimental class in University C followed the recommended program of the fitness training prescription generation system to teach physical education, and the cardiorespiratory fitness, muscular fitness, flexibility, and body composition of the experimental class changed significantly compared with that of the control class in the traditional teaching (P<0.05). 0.05).The outcome shows that the system which makes physical fitness training plans, that this paper develops by utilizing a genetic algorithm, has advantages for increasing the physical fitness of students.
