基于特征选择和模糊支持向量机的刀具磨损状态识别  

Tools wear state recognition based on feature selection and fuzzy support vector machine

作　　者：赵明利[1] 袁一鸣 宋士杰 聂立新[1] ZHAO Mingli;YUAN Yiming;SONG Shijie;NIE Lixin(School of Mechanical Engineering,Henan Polytechnic University,Jiaozuo 454000,CHN)

机构地区：[1]河南理工大学机械与动力工程学院,河南焦作454000

出　　处：《制造技术与机床》2020年第11期115-120,共6页Manufacturing Technology & Machine Tool

摘　　要：为了提高在不平衡性样本上的刀具磨损状态识别准确率,提出了一种基于特征选择和模糊支持向量机的刀具磨损状态识别方法。该方法通过离散二进制粒子群算法对特征进行选择,剔除冗余特征和无关特征,避免算法陷入维度灾难。设计了混合模糊隶属度函数,构建模糊支持向量机分类模型,采用粒子群算法优化模糊支持向量机关键参数,实现了刀具磨损状态识别。实验结果表明,在不平衡性样本情况下,基于特征选择和模糊支持向量机的刀具磨损状态识别方法具备良好的学习能力,具有较高的识别准确率。To improve the recognition accuracy of the tool wear state on unbalanced samples,a tool wear state recognition method based on feature selection and fuzzy support vector machine are proposed.The method selects features by discrete binary particle swarm optimization algorithm,eliminates redundant features and irrelevant features,and avoids the algorithm falling into the dimensional disaster.Then the hybrid fuzzy membership function is designed,and the fuzzy support vector machine classification model is constructed.The particle swarm optimization algorithm is used to optimize the key parameters of the fuzzy support vector machine to realize the tool wear state recognition.The experimental results show that in the case of unbalanced samples,the tool wear state recognition method based on feature selection and fuzzy support vector machine has good learning ability and high recognition accuracy.

关 键 词：特征选择 模糊支持向量机 刀具磨损 离散二进制粒子群算法 

分 类 号：TG71[金属学及工艺—刀具与模具] TP183[自动化与计算机技术—控制理论与控制工程]