目的 研究高速铣削参数对航空铸造钛合金Ti-6Al-4V表面质量的影响规律及交互作用，并基于高速铣削参数对表面质量和材料去除率进行优化。方法 采用Box-Behnken设计和二次回归正交实验法，建立高速铣削参数与表面粗糙度的显著不失拟回归模型，获得铣削参数影响表面粗糙度的显著性差异，挖掘高速铣削参数交互作用与表面粗糙度的关系;基于表面粗糙度回归模型及材料去除率，采用遗传算法(GA)，对高速铣削参数进行多目标优化。结果 铣削参数影响航空铸造钛合金Ti-6Al-4V试件表面粗糙度的显著性顺序为:切削深度>每齿进给量>切削宽度>主轴转速，其中切削宽度和主轴转速、每齿进给量和主轴转速的交互作用较为明显。利用遗传算法对铣削参数优化后，Ti-6Al-4V表面粗糙度较优化前提高44%，材料去除率提高70%，遗传算法优化后的试件表面粗糙度显著降低，表面刀路行距减小，纹理平均高度降低。结论 由实验验证可知，通过响应曲面建立表面粗糙度显著不失拟回归模型具有较高的预测精度，基于遗传算法优化获得的铣削参数可有效提高表面质量和切削效率，对保证航空铸造钛合金Ti-6Al-4V表面质量具有较好的指导意义。

The work aims to study the effect law and interaction of high-speed milling parameters on the surface quality of aviation casting titanium alloy Ti-6Al-4V, and optimize the surface quality and material removal rate based on high-speed milling parameters. Box-Behnken design and quadratic regression orthogonal test were adopted to establish a significant no-fail regression model of surface roughness and high-speed milling parameters, and then obtain significant differences in the effects of milling parameters on surface roughness. On this basis, the relationship between interaction of high-speed milling parameters and surface roughness was explored. Based on surface roughness regression model and material removal rate, Genetic Algorithm (GA) was used for multi-objective optimization of high-speed milling parameters. The order of significance of the milling parameters affecting the surface roughness of the aviation casting titanium alloy Ti-6Al-4V specimen was:cutting depth > feed per tooth > cutting width > spindle speed, in which the interaction of cutting width and spindle speed, feed per tooth and spindle speed was more obvious in interaction. The values of surface roughness and material removal rate were improved by 44% and 70%, respectively after the optimization of milling parameters by GA. Therefore, the surface roughness of the specimen optimized by GA was significantly reduced, the surface toolpath row spacing was narrowed, and the average height of the texture was lowered. The experimental results indicate that the significant no-fail regression model of surface roughness established by response surface has high prediction accuracy, and the milling parameters optimized by GA can effectively improve the surface quality and cutting efficiency, which is a good guideline for ensuring the surface quality of aviation casting titanium alloy Ti-6Al-4V.