To improve the anti-fatigue performance and the service life of parts made of 7050 aluminium alloy, the surface integrity of the workpiece treated by two-dimensional ultrasonic surface rolling (TDUSR) was comprehensively evaluated, and the optimization methods of the process parameters were proposed in this paper. The orthogonal experimental method was used to design an experiment program for TDUSR 7050 aluminium alloy. Based on the experimental results, the entropy weight method was used to make objectively weight for each quantifiable evaluation index. The technique for order preference by similarity to an ideal solution (TOPSIS) improved by the grey relational analysis (GRA) was presented to evaluate the surface integrity. Quadratic regression analysis was applied to establish prediction models for the surface roughness, microhardness, residual stress and the close degree. The optimal process parameters for the surface integrity were obtained using planning solution based on the prediction models. The results show that the static load and the feed rate have significant effects on the surface integrity, and their contributions are 60.52% and 28.05%, respectively. The maximal close degree obtained by planning solution is bigger than that obtained by TOPSIS improved by GRA. The validation test result proves that the overall evaluation method of TOPSIS improved by GRA combined with prediction modelling method by quadratic regression analysis can provide comparatively better process parameters and obtain more desirable process responses.

为了提高7050铝合金零件的抗疲劳性能和使用寿命，对二维超声表面轧制（TDUSR）处理的工件的表面完整性进行了综合评估，并提出了工艺参数的优化方法。在本文中。采用正交试验法设计了TDUSR 7050铝合金的实验程序。根据实验结果，采用熵权法对每个可量化评价指标进行客观权重确定。提出了一种通过与理想解决方案（TOPSIS）相似度来进行优先顺序排序的技术，该技术通过灰色关联分析（GRA）得到了改进，以评估表面完整性。应用二次回归分析来建立表面粗糙度，显微硬度，残余应力和紧密度。使用基于预测模型的规划解决方案获得了用于表面完整性的最佳工艺参数。结果表明，静载荷和进给速度对表面完整性有显着影响，其贡献分别为60.52％和28.05％。通过规划解获得的最大接近度大于通过GRA改进的TOPSIS获得的最大接近度。验证测试结果证明，通过GRA改进的TOPSIS总体评价方法与二次回归分析的预测建模方法相结合，可以提供相对较好的过程参数，并获得更理想的过程响应。使用基于预测模型的规划解决方案获得了用于表面完整性的最佳工艺参数。结果表明，静载荷和进给速度对表面完整性有显着影响，其贡献分别为60.52％和28.05％。通过规划解获得的最大接近度大于通过GRA改进的TOPSIS获得的最大接近度。验证测试结果证明，采用GRA改进的TOPSIS总体评价方法与二次回归分析的预测建模方法相结合，可提供相对较好的过程参数，并获得更理想的过程响应。使用基于预测模型的规划解决方案获得了用于表面完整性的最佳工艺参数。结果表明，静载荷和进给速度对表面完整性有显着影响，它们的贡献分别为60.52％和28.05％。通过规划解获得的最大接近度大于通过GRA改进的TOPSIS获得的最大接近度。验证测试结果证明，采用GRA改进的TOPSIS总体评价方法与二次回归分析的预测建模方法相结合，可以提供相对较好的过程参数，并获得更理想的过程响应。通过规划解获得的最大接近度大于通过GRA改进的TOPSIS获得的最大接近度。验证测试结果证明，采用GRA改进的TOPSIS总体评价方法与二次回归分析的预测建模方法相结合，可以提供相对较好的过程参数，并获得更理想的过程响应。通过规划解获得的最大接近度大于通过GRA改进的TOPSIS获得的最大接近度。验证测试结果证明，通过GRA改进的TOPSIS总体评价方法与二次回归分析的预测建模方法相结合，可以提供相对较好的过程参数，并获得更理想的过程响应。