Language is the bridge of communication between people. Different countries or regions have different social backgrounds, languages, and so on, so cultural differences must be paid attention to in the translation process. This paper focuses on exploring ways to improve students’ culturally adaptive translation ability during translation teaching. The particle swarm algorithm is employed to enhance the calculation model and create the analysis model for students’ culturally adaptive translation ability. The model is trained in Matlab software, and the correlation between the expected value and the measured value is measured by regression analysis to test the model’s performance. Subsequently, the model of this paper was applied to teaching practice, an experimental class was set up, a parallel control class was selected for reference, and a comparative analysis of the students’ culturally adaptive translation ability was conducted before and after the experiment in order to verify the effect of the application of the model of this paper on the improvement of the students’ culturally adaptive translation ability. The average score of students in the experimental class increased by 34.78% before and after the experiment, and the average score of students in the experimental class increased by 19.23% compared with that of students in the control class after the experiment. Compared to the pre-experiment, the increase in the scores of the students in the two classes for the culturally adaptive translation of long and difficult sentences was 116.67% and 47.62%, respectively.