This study explores the influencing factors of deep and meaningful learning in blended learning modes and their interrelationships. The sample comprises 397 college students from a university in Sichuan Province, selected through random sampling. Data collection utilized a questionnaire based on Bandura's ternary interaction theory, covering dimensions such as learners, helpers, environment, and interaction. Hypotheses were formulated based on literature research, and a survey was developed using established scales. Quantitative analysis was conducted using SPSS and AMOS software. The mean, standard deviation, variance, skewness, and kurtosis values were within reasonable ranges. The latent variables of the model exhibited sound convergent validity, with SFL ranging from 0.807 to 0.965, AVE from 0.697 to 0.946, and CR from 0.919 to 0.946. Model fit indices indicated acceptable fit (CMIN/DF: 2.303, NFI: 0.966, CFI: 0.980, RMSEA: 0.58, RMR: 0.008, PNFI: 0.789). Through path analysis, the study optimized the model, resulting in the final structural equation model (SEM). The findings suggest that (1) learner, environmental, and interaction factors positively influence deep and meaningful learning, while helper factors show a negative correlation; (2) learner, interaction, and helper factors mediate the environment's impact on deep and meaningful learning; and (3) environmental factors have the most significant influence on helper factors, followed by interaction and learner factors. Helpers play a significant role in enhancing deep understanding for learners. These insights guide effective strategies in promoting deep and meaningful learning in blended learning environments.