基于连续损伤力学（CDM）理论与支持向量机（SVM）模型，建立了一种疲劳寿命预测的新方法，旨在提高2014-T6铝合金材料疲劳寿命预测的准确性。首先，通过采用连续损伤力学模型及基于ABAQUS的UMAT子程序的二次开发，建立了一种用于预测2014-T6铝合金疲劳寿命的损伤力学有限元数值实现方法，并提出了基于粒子群算法的材料参数的标定方法。然后，为了进一步优化预测结果，利用支持向量机模型对基于损伤力学的疲劳寿命预测结果的误差进行训练，从而修正数值预测结果。通过将损伤力学有限元的预测结果、支持向量机模型修正后的预测结果与实验结果进行对比，发现采用支持向量机模型修正后的预测结果的精度较高，验证了所提方法的适用性，为2014-T6铝合金疲劳寿命预测提供了一个有效的解决方案，有望在实际工程应用中发挥重要作用。

Based on the Continuous Damage Mechanics （CDM） theory and Support Vector Machine （SVM） model， a novel fatigue life prediction method has been developed to improve the accuracy of fatigue life prediction for 2014-T6 aluminum alloy materials. Firstly， by adopting the continuous damage mechanics model and the secondary development of the Abaqus-based UMAT subroutine， a damage mechanics finite element numerical implementation method for predicting the fatigue life of 2014-T6 aluminum alloy is established， and a calibration method for material parameters based on the particle swarm optimization algorithm is proposed. Subsequently， to further optimize the prediction results， the SVM model is employed to train the errors of fatigue life prediction results based on damage mechanics， thereby correcting the numerical prediction results. By comparing the prediction results of damage mechanics finite element， the SVM model-corrected prediction results， and experimental results， it is found that the accuracy of the SVM model-corrected prediction results is higher， verifying the applicability of the proposed method. This provides an effective solution for the fatigue life prediction of 2014-T6 aluminum alloy and is expected to play a significant role in practical engineering applications.