Tool Condition Monitoring and Remaining Useful Life Prognostic Based on a Wireless Sensor in Dry Milling Operations
Tool breakage causes losses of surface polishing and dimensional accuracy for machined part, or possible damage to a workpiece or machine. Tool Condition Monitoring (TCM) is considerably vital in the manufacturing industry. In this paper, an indirect TCM approach is introduced with a wireless triaxi...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2016-05-01
|
Series: | Sensors |
Subjects: | |
Online Access: | http://www.mdpi.com/1424-8220/16/6/795 |
id |
doaj-f937b7429e614e94b2f5b13674df1c07 |
---|---|
record_format |
Article |
spelling |
doaj-f937b7429e614e94b2f5b13674df1c072020-11-25T01:00:42ZengMDPI AGSensors1424-82202016-05-0116679510.3390/s16060795s16060795Tool Condition Monitoring and Remaining Useful Life Prognostic Based on a Wireless Sensor in Dry Milling OperationsCunji Zhang0Xifan Yao1Jianming Zhang2Hong Jin3School of Mechanical and Automotive Engineering, South China University of Technology, Guangzhou 510640, ChinaSchool of Mechanical and Automotive Engineering, South China University of Technology, Guangzhou 510640, ChinaSchool of Mechanical and Automotive Engineering, South China University of Technology, Guangzhou 510640, ChinaCollege of Engineering, South China Agricultural University, Guangzhou 510642, ChinaTool breakage causes losses of surface polishing and dimensional accuracy for machined part, or possible damage to a workpiece or machine. Tool Condition Monitoring (TCM) is considerably vital in the manufacturing industry. In this paper, an indirect TCM approach is introduced with a wireless triaxial accelerometer. The vibrations in the three vertical directions (x, y and z) are acquired during milling operations, and the raw signals are de-noised by wavelet analysis. These features of de-noised signals are extracted in the time, frequency and time–frequency domains. The key features are selected based on Pearson’s Correlation Coefficient (PCC). The Neuro-Fuzzy Network (NFN) is adopted to predict the tool wear and Remaining Useful Life (RUL). In comparison with Back Propagation Neural Network (BPNN) and Radial Basis Function Network (RBFN), the results show that the NFN has the best performance in the prediction of tool wear and RUL.http://www.mdpi.com/1424-8220/16/6/795tool condition monitoring (TCM)remaining useful life (RUL)wireless sensorwavelet analysiswavelet packet transform (WPT)neuro-fuzzy network (NFN) |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Cunji Zhang Xifan Yao Jianming Zhang Hong Jin |
spellingShingle |
Cunji Zhang Xifan Yao Jianming Zhang Hong Jin Tool Condition Monitoring and Remaining Useful Life Prognostic Based on a Wireless Sensor in Dry Milling Operations Sensors tool condition monitoring (TCM) remaining useful life (RUL) wireless sensor wavelet analysis wavelet packet transform (WPT) neuro-fuzzy network (NFN) |
author_facet |
Cunji Zhang Xifan Yao Jianming Zhang Hong Jin |
author_sort |
Cunji Zhang |
title |
Tool Condition Monitoring and Remaining Useful Life Prognostic Based on a Wireless Sensor in Dry Milling Operations |
title_short |
Tool Condition Monitoring and Remaining Useful Life Prognostic Based on a Wireless Sensor in Dry Milling Operations |
title_full |
Tool Condition Monitoring and Remaining Useful Life Prognostic Based on a Wireless Sensor in Dry Milling Operations |
title_fullStr |
Tool Condition Monitoring and Remaining Useful Life Prognostic Based on a Wireless Sensor in Dry Milling Operations |
title_full_unstemmed |
Tool Condition Monitoring and Remaining Useful Life Prognostic Based on a Wireless Sensor in Dry Milling Operations |
title_sort |
tool condition monitoring and remaining useful life prognostic based on a wireless sensor in dry milling operations |
publisher |
MDPI AG |
series |
Sensors |
issn |
1424-8220 |
publishDate |
2016-05-01 |
description |
Tool breakage causes losses of surface polishing and dimensional accuracy for machined part, or possible damage to a workpiece or machine. Tool Condition Monitoring (TCM) is considerably vital in the manufacturing industry. In this paper, an indirect TCM approach is introduced with a wireless triaxial accelerometer. The vibrations in the three vertical directions (x, y and z) are acquired during milling operations, and the raw signals are de-noised by wavelet analysis. These features of de-noised signals are extracted in the time, frequency and time–frequency domains. The key features are selected based on Pearson’s Correlation Coefficient (PCC). The Neuro-Fuzzy Network (NFN) is adopted to predict the tool wear and Remaining Useful Life (RUL). In comparison with Back Propagation Neural Network (BPNN) and Radial Basis Function Network (RBFN), the results show that the NFN has the best performance in the prediction of tool wear and RUL. |
topic |
tool condition monitoring (TCM) remaining useful life (RUL) wireless sensor wavelet analysis wavelet packet transform (WPT) neuro-fuzzy network (NFN) |
url |
http://www.mdpi.com/1424-8220/16/6/795 |
work_keys_str_mv |
AT cunjizhang toolconditionmonitoringandremainingusefullifeprognosticbasedonawirelesssensorindrymillingoperations AT xifanyao toolconditionmonitoringandremainingusefullifeprognosticbasedonawirelesssensorindrymillingoperations AT jianmingzhang toolconditionmonitoringandremainingusefullifeprognosticbasedonawirelesssensorindrymillingoperations AT hongjin toolconditionmonitoringandremainingusefullifeprognosticbasedonawirelesssensorindrymillingoperations |
_version_ |
1725212430583201792 |