In this paper, fatigue tests were conducted on specimens of 2024 aluminum alloy welded joints before and after ultrasonic impact treatment (UIT). The test results show that after UIT, the fatigue life of the specimens significantly increased under identical stress levels (Δσ of 200MPa, 175MPa, 150MPa, and 125MPa), all reaching 10 7 cycles, indicating a substantial increase in the fatigue limit of the specimens.Building on these findings, a new model for studying fatigue life was employed, effectively integrating Continuous Damage Mechanics (CDM) theory with Artificial Neural Networks (ANN). Initially, a CDM model considering residual stresses was developed theoretically, and fatigue life was numerically calculated based on this model. More than 400 sets of data were collected to train the ANN; subsequently, predictions and validations of fatigue life were performed.The research results demonstrate that the proposed predictive model can accurately evaluate the effect of UIT on the fatigue life of 2024 aluminum alloy welded joint specimens, showcasing its effectiveness and applicability in predicting fatigue behavior.