摘要
针对批量钻削工序质量检测问题,采用声发射传感器采集工序加工过程中的声发射信号,提取其时域统计特征,构造工序过程信号的特征向量,根据密度带噪声的空间增量聚类算法(InDBSCAN)对工序过程中的声发射信号特征向量进行增量聚类,以分析批量工序质量。考虑到插入数据点在促成新类创建的同时可能引起已存在的不同类合并的情况,改进InDBSCAN算法。实验结果表明:改进的InDBSCAN算法使插入数据点的增量聚类更加合理,工序质量分布状况检测准确率达84.03%。
Aiming at monitor and analysis on batch drilling-quality,an acoustic emission sensor was used to collect the acoustic emission signal,extract statistic characteristics and then construct the signal characteristic vector.An improved incremental density based spatial clustering algorithm of time-domain applications with noise（InDBSCAN） was put forward to analyze the distribution law of batch drilling-quality indirectly.Take new data insertion into consideration.Because some of the original clusters could be remerged when the new cluster was created,and so the InDBSCAN algorithm was modified.The results show that the conclusion of incremental cluster analysis is more reasonable by the improved InDBSCAN algorithm and the detection accuracy of batch drilling-quality is up to 84.3%.
出处
《中南大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2012年第2期505-510,共6页
Journal of Central South University:Science and Technology
基金
湖南省高校创新平台开放基金资助项目(10K063)
教育部留学回国人员科研启动基金资助项目(2009-1590)
湖南省自然科学基金省市联合基金资助项目(10JJ9005)
关键词
批量钻削
工序质量
特征向量
增量聚类
层次分析法
batch drilling
process quality
characteristic vector
incremental clustering
analytic hierarchy process
