基于遗传支持向量机的车床刀具寿命预测  

Prediction of cutting-tool life based on support vector machine and genetic algorithm

作　　者：杨路[1] 

机构地区：[1]江南大学信息工程学院,江苏无锡214064

出　　处：《微计算机信息》2010年第34期257-258,共2页Control & Automation

摘　　要：刀具寿命预测对于换刀决策支持有着重要意义。本文结合支持向量机与遗传算法,提出一种刀具寿命预测新方法,以实现刀具的准确预测。利用遗传算法确定SVM中的训练参数,以得到优化的SVM预测模型,并利用SVM在小样本、非线性中优越的预测性能对刀具寿命进行预测。试验结果表明,结合遗传算法与支持向量机预测方法对刀具寿命进行预测具有很好的预测精度。Prediction of cutting-tool life is significant to decision support of tool changing.In the paper,the new method combining support vector machine and genetic algorithm is proposed to forecast cutting-tool life in order to realize the exact forecasting of cutting-tool life.Genetic algorithm（GA） is used to determine training parameters of support vector machine in this model,which can gain optimized SVM forecasting model.And cutting-tool life is forecasted by utilizing excellent forecasting performance in small sample and nonlinear of support vector machine.The experimental results indicate that the proposed GA-SVM model can achieve great accuracy in cutting-tool life prediction.

关 键 词：刀具寿命 支持向量机 遗传算法 预测 

分 类 号：TG713.1[金属学及工艺—刀具与模具]