钛合金立铣表面粗糙度预测新方法

为实现加工前对表面粗糙度的预测,建立高精度的表面粗糙度预测模型至关重要.针对钛合金立铣表面粗糙度的特点及传统预测方法的不足,提出了表面粗糙度预测新方法.分别用递推最小二乘算法、基本蚁群算法与混合蚁群算法训练模糊系统,混合蚁群算法的收敛效果优于递推最小二乘算法和基本蚁群算法.通过回归分析建立了表面粗糙度的两种经验公式.对各方法所得模型进行测试,结果表明混合蚁群算法训练模糊系统的预测效果优于其他方法,用混合蚁群算法训练的模糊系统进行表面粗糙度预测是可行的.

A Novel Method for Predicting Surface Roughness in End Milling of Titanium Alloy

The model of predicting surface roughness with great accuracy is important for predicting sur- face roughness before cutting. With regard to characteristics of the surface roughness of end milling of ti- tanium alloy and disadvantages of traditional methods, a novel method for predicting surface roughness is developed. Fuzzy system is trained by recursive least-square algorithm, basic ant colony algorithm and hybrid ant colony algorithm respectively. The convergence error of hybrid ant colony algorithm is small compared with that of recursive least square algorithm and basic ant colony algorithm. Two empiric formu- las of surface roughness are developed by regression analysis. The models obtained from the methods mentioned above are tested. Results show that the accuracy of the fuzzy system trained by hybrid ant colo- ny algorithm is the highest and that the hybrid ant colony algorithm-based fuzzy system can be reliably ap- plied to predict surface roughness.