The rise of intelligent sports has led to the development of the sports business, and at the same time, it also provides reference methods and ideas for the innovation and development of sports teaching. This paper proposes a personalized training program for physical education based on intelligent algorithms, by constructing a sports prescription generation model based on the artificial raindrop algorithm to meet students’ needs for personalized sports. The study proposes an exercise prescription generation model based on the artificial raindrop algorithm, which optimizes the exercise training parameters by improving the artificial raindrop operator, and then generates a decision scheme for exercise prescription. After applying the method proposed in this paper to specific physical education teaching, it was found that the average lung capacity of male students in the experimental group was improved by 128.05 ml, which is a highly significant difference compared with the traditional sports training methods. Overall, the exercise prescription proposed in this paper can effectively improve the physical fitness level of secondary school students, and its effect is better than that of traditional physical education teaching methods.