In this era of deep integration of knowledge explosion and globalization, the importance of higher education, as the cradle of cultivating the pillars of the future society, is self-evident. As a bridge connecting the world culture, college English learning plays an even important role. However, the traditional English teaching model is often limited by the “one-size-fits-all” teaching method and rigid curriculum design, which leads to low learning efficiency and frustration of students’ enthusiasm, and it is difficult to fully meet the personalized learning needs of different students. Faced with this challenge, we urgently need an innovative teaching approach to break the traditional framework, improve learning efficiency, as an intelligent optimization technology simulating the behavior of natural biological groups, has brought new enlightenment to the field of education with its unique optimization mechanism and non-linear dynamic adjustment ability. This paper introduces the PSO algorithm into the exploration of college English learning mode, aiming to dynamically adjust the learning path by simulating group wisdom, and tailor the best learning scheme for each student. The MATLAB simulation results show that under certain evaluation criteria, the PSO algorithm is superior to the traditional neural network algorithm in terms of learning model research accuracy and learning model research influencing factor time.