Failure Prediction for CNC Machine Tool Feed System Based on Hidden Markov Models

摘要：

　　The reliability of CNC machine tool feed system plays an important role in the normal operations of machine tools.This paper applies the hidden Markov model to failure prediction for machine tools feed system.Firstly,we collect the vibration signals of the machine tool feed system in both healthy and faulty conditions by sensors.We extract the features of the signals to acquire observation sequences.Then we use the observation sequences to train the models and form the model library.Finally,the trained model library is used to predict the failures of the feed system.The experiments illustrate that the proposed approach is effective in feed system failure prediction.

关键词： Feed System Hidden Markov Model Failure Prediction

作者: Bei ZHANG Huifen WANG Tingyu LIU

作者单位: School of Mechanical Engineering,Nanjing University of Science and Technology,Nanjing,Jiangsu 210094,P.R.China

会议类型: 国际会议

会议名称: The 5th International Conference on Mechanical Engineering and Mechanics(第五届国际机械工程与力学会议)

会议地点: 江苏扬州

会议语种:英文

页码: 497-503

在线出版日期: 2014-08-21（万方平台首次上网日期，不代表论文的发表时间）