侧铣加工刀具回转轮廓误差预测技术研究

Research on Prediction Technology for Tool Rotation Profile Errors in Flank Milling

在侧铣加工中,刀具磨损和变形引起的刀具回转轮廓误差在实际加工前难以准确预测。提出一种工件形状刀具轮廓映射的辨识试验方法来获取加工过程刀具回转轮廓误差,并通过多因素正交试验获取了不同工况下刀具回转轮廓误差数据库。基于误差数据,采用最小二乘支持向量机(LS-SVM)技术建立了刀具回转轮廓误差预测模型。运用遗传算法优化对所提模型有重要影响的核函数参数和错误惩罚因子,建立了基于遗传算法优化的最小二乘支持向量机(GA-LS-SVM)模型,并与未经遗传算法优化的LS-SVM模型进行了对比,试验结果表明,GA-LS-SVM预测模型能更好地适用于刀具回转轮廓误差预测。

In the flank milling processes,it is difficult to accurately predict the tool rotation profile errors caused by tool wears and deformations before actual machining.A identification test method of workpiece shape-tool profile mapping was proposed to obtain the tool rotation profile errors and the error data under different working conditions were obtained through the multi-factor orthogonal tests.A prediction model for tool rotation profile error was established by the LS-SVM technology based on the error data.The genetic algorithm(GA)was used to optimize the model parameters including kernel function parameters and error warning factors,which were very important to the proposed model.A LS-SVM model was established based on GA optimization(GA-LS-SVM),which was compared with a LS-SVM model without GA optimization.The testing results show that the GA-LS-SVM prediction model is more suitable for tool rotation profile error prediction.