There is an issue with erroneous models, despite the fact that student models play a significant part in the online learning system for English grammar. The online English grammar learning system’s model research issue is intractable and yields unsatisfactory results when solved using the conventional ant colony approach. Hence, this work evaluates based on evolutionary algorithms and offers further research to support this model. In order to minimize interference with the student model’s research and execution, we first use chromosomal and fitness theory to identify the elements that will have an impact, and then we split the indicators according to the needs of the student model. Next, the student model’s research and implementation plan is developed utilizing chromosomal and fitness theory for the genetic algorithm. The student model’s research and implementation outcomes are then thoroughly evaluated. The results show that the intelligent analysis method can improve the learning system and English learning in this paper, with the improvement degree reaching more than 80%, and reducing the corresponding changes. The reduction rate of change amplitude is 10%. Therefore, it can promote the optimization of English grammar and online learning.