An enhanced ant colony algorithm-oriented aerobic exercise approach was proposed to study the association between youth aerobic exercise and obesity reduction. To begin, changes in body shape, weight, BMI, body fat, body circumference, and other indicators of obese adolescents before and after aerobic exercise were used as the initial pheromone distribution matrix, and the random evolution factor and evolutionary drift threshold were introduced to establish the target function of reducing obesity in adolescents caused by aerobic exercise. The particle algorithm was proposed to construct the best model of aerobic exercise for teenage obesity reduction, and the constraint conditions of the association between aerobic exercise and adolescent obesity reduction. The average modelling error of this method is about 0.053 percent, while the traditional method's average error is about 0.186 percent, indicating that this method can control the error within a reasonable range, and it has been proven that the improved ant colony algorithm can have a good correlation with the method of aerobic exercise.
