铣削刀具磨损状态监测系统关键技术研究

发布时间：2018-04-03 13:44

本文选题：铣削刀具　切入点：刀具磨损　出处：《哈尔滨理工大学》2016年硕士论文

【摘要】：随着工业智能化的提出,数控加工技术正朝着智能化的方向发展。刀具磨损状态监测技术在数控加工智能化发展中占据重要的地位,在加工过程中,精确及时地监测到刀具磨损状态并且更换失效的刀具,不仅有助于提高生产效率,还可以提高刀具的使用寿命。因此,刀具磨损状态监测系统是现代数控加工智能化发展迫切需要的,其关键技术的研究具有十分重要的意义。本文针对刀具磨损的形成机理进行了详尽的分析,根据分析制定刀具磨钝标准,考虑不同磨损状态下对监测信号的影响,合理选择监测方法,建立铣刀磨损状态监测系统的信号采集平台,制定信号采集方案并且采集信号数据。针对上述所采集的数据进行特征提取,建立特征向量与铣削刀具磨损的之间的关系,针对特征向量维数高的特点,提出了改进的多分类支持向量机递归特征消去的特征选择方法,运用测试样本进行验证,该方法优于其他特征选择方法,为后续的状态识别,提供维数低和对不同铣刀磨损状态比较敏感的特征向量。采用万有引力优化最小二乘支持向量机的铣刀磨损状态识别的方法,利用选择后的特征向量进行铣刀磨损状态的识别。通过对比,该方法的识别精度较高。本文最后结合VC++软件和Matlab软件的优点,开发铣削刀具磨损状态监测系统,可以实现铣削加工中刀具磨损状态的快速识别,通过识别结果显示界面,用户可以直观地判断刀具磨损的状况,进而判断是否更换刀具,减少停机时间,提高了加工效率。通过对铣削刀具磨损状态监测系统关键技术的研究,提高了刀具的利用率,避免了报废工件的产生,降低了生产过程中资源的大量浪费,提高了生产效率。为数控加工智能化的发展奠定了基础。
[Abstract]:With the advance of industrial intelligence, NC machining technology is developing towards the direction of intelligentization.Tool wear monitoring technology plays an important role in the intelligent development of NC machining. In the process of machining, it is not only helpful to improve the production efficiency, but also to monitor the tool wear state accurately and timely and to replace the invalid tool in the process of machining.It can also improve the service life of cutting tools.Therefore, tool wear monitoring system is an urgent need for the intelligent development of modern NC machining, and the research of its key technology is of great significance.In this paper, the formation mechanism of tool wear is analyzed in detail. According to the analysis, the tool wear bluntness standard is formulated, the influence of different wear states on monitoring signal is considered, and the monitoring method is reasonably selected.The signal acquisition platform of milling cutter wear monitoring system is established, and the signal acquisition scheme and signal data collection are made.Based on the above data, the relationship between feature vector and milling tool wear is established, and the feature vector dimension is high.An improved feature selection method based on recurrent feature elimination of multi-classification support vector machines is proposed, which is verified by test samples. This method is superior to other feature selection methods and is used for subsequent state recognition.Provides feature vectors with low dimension and sensitivity to different milling cutter wear states.The wear state of milling cutter is identified by using the universal gravity optimization least squares support vector machine (LS-SVM) and the selected eigenvector.By comparison, the recognition accuracy of this method is high.Finally, combining the advantages of VC software and Matlab software, a tool wear monitoring system is developed, which can realize the rapid recognition of tool wear state in milling machining, and display the interface through the recognition results.The user can directly judge the condition of tool wear, and then judge whether to change the tool, reduce the downtime and improve the machining efficiency.Through the research on the key technology of the monitoring system of milling tool wear status, the utilization rate of the tool is improved, the production of scrapped workpiece is avoided, the waste of resources in the production process is reduced, and the production efficiency is improved.It lays a foundation for the development of intelligent NC machining.
【学位授予单位】：哈尔滨理工大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TG54;TG714

【相似文献】