跳到主要內容

臺灣博碩士論文加值系統

(44.200.140.218) 您好!臺灣時間:2024/07/19 02:51
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

: 
twitterline
研究生:楊勝雄
研究生(外文):YANG SHENG-HSIUNG
論文名稱:半導體設備的預防性異常檢測技術探討 -以時規皮帶張力異常偵測為例
論文名稱(外文):Exploration of Proactive Anomaly Detection Techniques for Semiconductor Equipment: A Case Study on Abnormal Detection of Timing Belt Tension
指導教授:李俊宏李俊宏引用關係
指導教授(外文):LEE CHUNG-HONG
口試委員:張道行楊新章李俊宏
口試委員(外文):CHANG TAO-HSINGYANG HSIN-CHANGLEE CHUNG-HONG
口試日期:2024-01-17
學位類別:碩士
校院名稱:國立高雄科技大學
系所名稱:電機工程系
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2024
畢業學年度:112
語文別:中文
論文頁數:104
中文關鍵詞:劣化診斷皮帶張力物聯網感測器機器學習
外文關鍵詞:deterioration diagnosisbelt tensionIoT sensorsmachine learning
相關次數:
  • 被引用被引用:0
  • 點閱點閱:110
  • 評分評分:
  • 下載下載:29
  • 收藏至我的研究室書目清單書目收藏:0
在半導體設備上因發生了未能及時偵測出的負載異常,導致皮帶輪傳動軸斷
裂,使得生產作業被迫停止,造成產線的嚴重損失。為了探討馬達驅動器的劣化
診斷功能失效的原因,本研究以皮帶張力異常為檢測目標,以實驗重現劣化診斷
功能的辨識結果,嘗試找出診斷失敗的原因。為了評估其他的檢測方案,將結合
機器學習的物聯網感測裝置納入研究。
劣化診斷是伺服驅動器所附加的一種預測性檢測功能,以驅動器內部參數值
設定作為劣化診斷門檻的判定依據,實驗結果表明,門檻值會受馬達轉速限制而
影響劣化診斷的判定。為補強劣化診斷功能的不足,本研究另提出機器學習結合
物聯網感測器的預防性檢測方法,是以振動信號為異常樣本來進行模型訓練,經
過訓練後的模型,能夠依據振動信號來進行異常狀態的判別,並輸出相應的分類
結果,最後再以多數票方式的統計方式作為判定依據,並且可以在不同轉速下得
到良好的辨識結果,若能將驅動器的劣化診斷功能整合機器學習的物聯網感測器,
並作為一種新的異常檢測方法,則可以彌補劣化診斷功能在速度變化下的盲點。
本研究提出將預測性(Pridictive)的故障徵兆檢測再結合預防性(Preventive)的
故障現況識別的方法,以提高對皮帶張力異常的檢測能力,並將此方法定義為主
動式異常檢測(Proactive Anomaly Detection)。
Due to an undetected load anomaly on semiconductor equipment, the drive shaft of the belt pulley fractured, causing a forced halt in production and resulting in severe losses in the production line. In order to investigate the reasons for the failure of the degradation diagnosis function of the motor drive, this study focuses on abnormal belt tension as the detection target. Through experiments replicating the identification results of the degraded diagnosis function, an attempt is made to identify the reasons for the diagnostic failure.
To evaluate alternative detection solutions, IoT sensing devices incorporating machine learning will be integrated into the research. Degradation diagnosis is a predictive detection function attached to servo drives, using set internal parameter values as the basis for degradation diagnosis threshold determination. Experimental results indicate that the threshold value is influenced by motor speed limitations, affecting the determination of degradation diagnosis.
To address the limitations of the degradation diagnosis function, this study proposes a preventive detection method combining machine learning with IoT sensors. The method involves training models using vibration signals as abnormal samples. The trained models can discern abnormal states based on vibration signals, output corresponding classification results, and use a majority voting statistical approach as the basis for determination. This approach yields good recognition results at different speeds. If the degradation diagnosis function of the drive can be integrated with machine learning IoT sensors as a novel anomaly detection method, it can compensate for blind spots in degradation diagnosis functions under speed variations.
This study proposes the combination of predictive fault symptom detection with
preventive fault current state identification to enhance the detection capability of abnormal belt tension. This method is defined as Proactive Anomaly Detection.
目錄
摘要 ................................................................................................................................... i
ABSTRACT ..................................................................................................................... ii
誌謝 ................................................................................................................................. iv
表目錄 ............................................................................................................................ vii
圖目錄 ........................................................................................................................... viii
第一章、 緒論 ........................................................................................................... 1
1.1 前言 .....................................................................................................................1
1.2 研究背景 .............................................................................................................2
1.3 研究動機 .............................................................................................................2
1.4 研究目的 .............................................................................................................4
1.5 論文架構 .............................................................................................................5
第二章、 文獻回顧與技術探討 ............................................................................... 6
2.1 預防性維護的應用 .............................................................................................7
2.2 物聯網與機器學習 .............................................................................................8
2.3 故障診斷與維護策略 .........................................................................................9
2.4 機器學習模型 ...................................................................................................10
2.5 相關技術探討 ...................................................................................................13
2.5.1 特徵生成方法-MFCC ............................................................................13
2.5.2 數據異常過濾法-Isolation Forest Filter ................................................16
第三章、 實驗架構與實驗方法 ............................................................................. 19
3.1 實驗架構 ...........................................................................................................19
3.1.1 實驗硬體 ................................................................................................21
3.1.2 伺服馬達工具軟體 ................................................................................24
3.1.3 IoT 感測模組..........................................................................................25
3.1.4 振動特徵波型 ........................................................................................27
3.2 伺服馬達調整 ...................................................................................................29
3.2.1 伺服驅動器參數調整 ............................................................................30
3.2.2 伺服馬達速度設定 ................................................................................32
3.2.3 伺服驅動器劣化診斷功能 ....................................................................34
3.3 機構異常檢出對策 ...........................................................................................39
3.3.1 時規皮帶磨損與張力 ............................................................................39
3.3.2 劣化診斷功能 ........................................................................................44
3.3.3 機器學習的 IoT 感測模組 ....................................................................46
3.4 實驗方法 ...........................................................................................................47
3.4.1 皮帶張力調整與限制 ............................................................................47
3.4.2 伺服劣化診斷功能實驗 ........................................................................51
3.4.3 機器學習的異常類別定義 ....................................................................54
3.4.4 振動數據蒐集與特徵標記 ....................................................................54
第四章、 實驗結果 ................................................................................................. 61
4.1 劣化診斷實驗 ...................................................................................................61
4.2 IoT 感測模組實驗 ...........................................................................................63
4.2.1 模型建立與評估 ....................................................................................63
4.2.2 模型韌體燒錄與部署 ............................................................................68
4.2.3 分類輸出結果的統計方法 ....................................................................78
第五章、 結論與未來研究 ..................................................................................... 79
5.1 結論 ...................................................................................................................79
5.2 未來研究 ...........................................................................................................80
參考文獻 ......................................................................................................................... 81
附錄一 ............................................................................................................................. 88

表目錄
表 1、滾珠螺桿滑台使用部件 ..................................................................................... 21
表 2、電腦硬體與工具軟體一覽表 ............................................................................. 22
表 3、運動條件設定 ..................................................................................................... 33
表 4、劣化診斷的門檻設定值 ..................................................................................... 36
表 5、劣化診斷項目參數表 [40] ................................................................................. 37
表 6、啟用劣化診斷功能必要項目參數 [40] ............................................................. 38
表 7、皮帶損壞原因與對策 ......................................................................................... 40
表 8、皮帶狀態與對應的張力值 ................................................................................. 49
表 9、伺服馬達運動模式的速度條件 ......................................................................... 51
表 10、劣化診斷時間與轉速參數的設定 ................................................................... 52
表 11、特徵類別標記說明 ........................................................................................... 54
表 12、用於訓練的標籤組與樣本數據量 ................................................................... 60
表 13、各張力與轉速的劣化診斷參數值 ................................................................... 62
表 14、張力 0N 辨識結果 ............................................................................................ 70
表 15、張力 30N 辨識結果 .......................................................................................... 71
表 16、張力 60N 辨識結果 .......................................................................................... 72
表 17、張力 120N 辨識結果 ........................................................................................ 73
表 18、靜止狀態與張力 120N 辨識結果 .................................................................... 74
表 19、運動速度 45kpps 之辨識結果 .......................................................................... 75
表 20、運動速度 65kpps 之辨識結果 .......................................................................... 76

圖目錄
圖 1、皮帶輪傳動軸斷裂 ............................................................................................... 3
圖 2、決策樹(左)與隨機森林(右) ................................................................................. 11
圖 3、線性可分 SVM ................................................................................................... 12
圖 4、非線性可分 SVM ............................................................................................... 12
圖 5、線性頻率與梅爾頻率特性曲線 ......................................................................... 13
圖 6、驅動器劣化診斷實驗架構 ................................................................................. 19
圖 7、IoT 感測器實驗架構 ........................................................................................... 20
圖 8、伺服螺桿滑台外觀 ............................................................................................. 23
圖 9、伺服馬達驅動時規皮帶 ..................................................................................... 23
圖 10、伺服工具軟體操作介面 ................................................................................... 24
圖 11、SeneorTile.box 內部電路板 .............................................................................. 25
圖 12、資料擷取韌體 ................................................................................................... 25
圖 13、韌體燒寫裝置連接 ........................................................................................... 26
圖 14、ST Cube Programmer 燒錄完成 ....................................................................... 26
圖 15、加速度計波形 ................................................................................................... 27
圖 16、Gyorscope wave form ........................................................................................ 28
圖 17、伺服驅動器增益控制架構 [40] ....................................................................... 29
圖 18、與驅動器連線的訊息對話框 ........................................................................... 30
圖 19、自動增益調整畫面 ........................................................................................... 31
圖 20、正常皮帶張力下的運動波形曲線圖 ............................................................... 32
圖 21、伺服馬達往返運轉的波形 ............................................................................... 34
圖 22、劣化診斷設定畫面 ........................................................................................... 35
圖 23、皮帶張力計算示意圖 ....................................................................................... 42
圖 24、發生轉矩指令平均值的劣化診斷警告情況 [40] ........................................... 45
圖 25、手動方式振動特徵標記 ................................................................................... 46
圖 26、皮帶張力計參數設定 ....................................................................................... 47
圖 27、IoT 感測模組的安裝位置 ................................................................................. 48
圖 28、IoT 振動感測模組固定方式 ............................................................................. 49
圖 29、劣化診斷發生警告 ........................................................................................... 53
圖 30、1-Normal 的振動波形 ....................................................................................... 55
圖 31、2-TooTight 的振動波形 .................................................................................... 56
圖 32、3-Loose 的振動波形 ......................................................................................... 56
圖 33、4-TooLoose 的波形 ........................................................................................... 57
圖 34、5-Stopping 的振動波形 ..................................................................................... 58
圖 35、Normal 波形的特徵標記 .................................................................................. 58
圖 36、Stopping 狀態的 Labeling ................................................................................. 59
圖 37、分類模型-1 混淆矩陣 ....................................................................................... 64
圖 38、分類模型-2 混淆矩陣 ....................................................................................... 64
圖 39、分類模型-3 混淆矩陣 ....................................................................................... 65
圖 40、分類模型-4 混淆矩陣 ....................................................................................... 66
圖 41、Feature Statistics Table ...................................................................................... 67
圖 42、即時辨識結果的輸出顯示 ............................................................................... 68
圖 43、馬達往返運動波形示意 ................................................................................... 69
圖 44、加速度計 ???? 通道與陀螺儀 ???? 通道波型示意 ....................................................... 69
參考文獻
[1] Zonta, T., et al., 2020, "Predictive maintenance in the Industry 4.0: A systematic
literature review", Computers & Industrial Engineering,Vol. 150, pp. 106889.
https://doi.org/10.1016/j.cie.2020.106889
[2] Sun, B., et al., 2017, "Perspective for equipment automation in process industries",
IFAC-PapersOnLine,Vol. 50(2), pp. 65-70.
https://doi.org/10.1016/j.ifacol.2017.12.012
[3] 林柏辰, 2017, 基於伺服驅動器輸出資料以模糊方法估測軸承損壞程度, 國立臺
北科技大學, 碩士論文. https://hdl.handle.net/11296/9hutkq
[4] 林晉枝, 2020, 半導體製程真空幫浦故障預測, 國立臺北科技大學, 碩士論文.
https://hdl.handle.net/11296/3cun25
[5] 彭善謙, 2004, 綜合振動信號於馬達故障診斷, 中原大學, 碩士論文.
https://hdl.handle.net/11296/be5559
[6] 陳景裕, 2020, 工具機故障預測與健康管理模式建構, 國立虎尾科技大學, 碩士論
文. https://hdl.handle.net/11296/g4j9a9
[7] Sharma, A. K., et al., 2021, "A study of trends and industrial prospects of Industry
4.0", Materials Today: Proceedings,Vol. 47, pp. 2364-2369.
https://doi.org/10.1016/j.matpr.2021.04.321
[8] Civerchia, F., et al., 2017, "Industrial Internet of Things monitoring solution for
advanced predictive maintenance applications", Journal of Industrial Information
Integration,Vol. 7, pp. 4-12. https://doi.org/10.1016/j.jii.2017.02.003
[9] Karmakar, A., et al., 2019, "Industrial internet of things: a review", 2019
international conference on opto-electronics and applied optics (optronix), pp. 1-6,
IEEE. https://doi.org/10.1109/OPTRONIX.2019.8862436
[10] Khan, W. Z., et al., 2020, "Industrial internet of things: Recent advances, enabling
technologies and open challenges", Computers & Electrical Engineering,Vol. 81, pp.
106522. https://doi.org/10.1016/j.compeleceng.2019.106522
[11] De Vita, F., et al., 2020, "On the use of a full stack hardware/software infrastructure
for sensor data fusion and fault prediction in industry 4.0", Pattern Recognition
Letters,Vol. 138, pp. 30-37. https://doi.org/10.1016/j.patrec.2020.06.028
[12] Reinfurt, L., et al., 2017, "Applying IoT patterns to smart factory systems",
Advanced Summer School on Service Oriented Computing, Summer SOC,Vol. 66.
https://www.iaas.uni-stuttgart.de/publications/INPROC-2017-64-Applying-IoT-
Patterns-to-Smart-Factory-Systems.pdf
[13] Chen, B., & Wan, J., 2019, "Emerging trends of ml-based intelligent services for
industrial internet of things (iiot)", 2019 Computing, Communications and IoT
Applications (ComComAp), pp. 135-139.
https://doi.org/10.1109/ComComAp46287.2019.9018815
[14] Janssens, O., et al., 2016, "Convolutional Neural Network Based Fault Detection
for Rotating Machinery", Journal of Sound and Vibration,Vol. 377, pp. 331-345.
https://doi.org/10.1016/j.jsv.2016.05.027
[15] Wu, W., et al., 2018, "A visual analytics approach for equipment condition
monitoring in smart factories of process industry", 2018 IEEE Pacific Visualization Symposium (PacificVis), pp. 140-149, IEEE.
https://doi.org/10.1109/PacificVis.2018.00026
[16] Jardine, A. K. S., et al., 2006, "A review on machinery diagnostics and prognostics
implementing condition-based maintenance", Mechanical systems and signal
processing,Vol. 20(7), pp. 1483-1510. https://doi.org/10.1016/j.ymssp.2005.09.012
[17] Li, H., & Ma, X., 2006, "Investigation on Fault Diagnosis System Based on Time
Spatial Information Fusion Theory", 2006 6th World Congress on Intelligent Control
and Automation, Vol. 2, pp. 5595-5599, IEEE.
https://doi.org/10.1109/WCICA.2006.1714145
[18] Dsouza, J., & Velan, S., 2019, "Preventive Maintenance for Fault Detection in
Transfer Nodes using Machine Learning", 2019 International Conference on
Computational Intelligence and Knowledge Economy (ICCIKE), pp. 401-404, IEEE.
https://doi.org/10.1109/ICCIKE47802.2019.9004230
[19] Christou, I. T., et al., 2020, "End-to-End Industrial IoT Platform for Actionable
Predictive Maintenance", IFAC-PapersOnLine,Vol. 53(3), pp. 173-178.
https://doi.org/10.1016/j.ifacol.2020.11.028
[20] Calabrese, M., et al., 2020, "SOPHIA: An Event-Based IoT and Machine Learning
Architecture for Predictive Maintenance in Industry 4.0", Information,Vol. 11(4), pp.
202. https://doi.org/10.3390/info11040202
[21] Dilda, V., et al., 2017, "Manufacturing: Analytics unleashes productivity and
profitability", Report, McKinsey & Company.
https://www.mckinsey.com/~/media/McKinsey/Business%20Functions/Operations/Our%20Insights/Manufacturing%20Analytics%20unleashes%20productivity%20and
%20profitability/Manufacturing-analytics-unleashes-productivity-and-
profitability.pdf
[22] Nasteski, V., 2017, "An overview of the supervised machine learning methods",
Horizons. b,Vol. 4, pp. 51-62. https://www.researchgate.net/profile/Vladimir-
Nasteski/publication/328146111_An_overview_of_the_supervised_machine_learnin
g_methods/links/5c1025194585157ac1bba147/An-overview-of-the-supervised-
machine-learning-methods.pdf
[23] Cover, T., & Hart, P., 1967, "Nearest neighbor pattern classification", IEEE
transactions on information theory,Vol. 13(1), pp. 21-27.
https://doi.org/10.1109/TIT.1967.1053964
[24] Guo, G., et al., 2003, "KNN model-based approach in classification", On The
Move to Meaningful Internet Systems 2003: CoopIS, DOA, and ODBASE: OTM
Confederated International Conferences, CoopIS, DOA, and ODBASE 2003, Catania,
Sicily, Italy, November 3-7, 2003. Proceedings, pp. 986-996, Springer.
https://doi.org/10.1007/978-3-540-39964-3_62
[25] Wei, C.-C., et al., 2013, "K-nn based neuro-fuzzy system for time series prediction",
2013 14th ACIS International Conference on Software Engineering, Artificial
Intelligence, Networking and Parallel/Distributed Computing, pp. 569-574, IEEE.
https://doi.org/10.1109/SNPD.2013.68
[26] Navada, A., et al., 2011, "Overview of use of decision tree algorithms in machine
learning", 2011 IEEE control and system graduate research colloquium, pp. 37-42,
IEEE. https://doi.org/10.1109/ICSGRC.2011.5991826
[27] Loh, W. Y., 2011, "Classification and regression trees", Wiley interdisciplinary
reviews: data mining and knowledge discovery,Vol. 1(1), pp. 14-23.
https://doi.org/10.1002/widm.8
[28] Biau, G., & Scornet, E., 2016, "A random forest guided tour", Test,Vol. 25, pp.
197-227. https://doi.org/10.1007/s11222-016-9646-1
[29] Puggini, L., et al., 2015, "Fault detection using random forest similarity distance",
IFAC-PapersOnLine,Vol. 48(21), pp. 583-588.
https://doi.org/10.1016/j.ifacol.2015.09.589
[30] Benkedjouh, T., et al., 2018, "Gearbox fault diagnosis based on mel-frequency
cepstral coefficients and support vector machine", IFIP International Conference on
Computational Intelligence and Its Applications, pp. 220-231, Springer.
https://doi.org/10.1007/978-3-319-89743-1_20
[31] Platt, J., 1999, "Probabilistic outputs for support vector machines and comparisons
to regularized likelihood methods", Advances in large margin classifiers,Vol. 10(3),
pp. 61-74. https://www.researchgate.net/profile/John-Platt-
2/publication/2594015_Probabilistic_Outputs_for_Support_Vector_Machines_and_C
omparisons_to_Regularized_Likelihood_Methods/links/004635154cff5262d6000000
/Probabilistic-Outputs-for-Support-Vector-Machines-and-Comparisons-to-
Regularized-Likelihood-Methods.pdf
[32] Cortes, C., & Vapnik, V., 1995, "Support-vector networks", Machine learning,Vol.
20, pp. 273-297. https://doi.org/10.1007/BF00994018
[33] Davis, S., & Mermelstein, P., 1980, "Comparison of parametric representations for
monosyllabic word recognition in continuously spoken sentences", IEEE transactions
on acoustics, speech, and signal processing,Vol. 28(4), pp. 357-366.
https://doi.org/10.1109/TASSP.1980.1163420
[34] Ittichaichareon, C., et al., 2012, "Speech recognition using MFCC", International
conference on computer graphics, simulation and modeling, Vol. 9,
https://www.researchgate.net/profile/Sloveby-
Suksri/publication/281446199_Speech_Recognition_using_MFCC/links/55e7e61d08
aeb6516262ec4c/Speech-Recognition-using-MFCC.pdf
[35] Akpudo, U. E., & Hur, J.-W., 2021, "A cost-efficient MFCC-based fault detection
and isolation technology for electromagnetic pumps", Electronics,Vol. 10(4), pp. 439.
https://doi.org/10.3390/electronics10040439
[36] Sousa, R., et al., 2019, "Robust cepstral-based features for anomaly detection in
ball bearings", The International Journal of Advanced Manufacturing Technology,Vol.
103, pp. 2377-2390. https://doi.org/10.1007/s00170-019-03597-2
[37] Breunig, M. M., et al., 2000, "LOF: identifying density-based local outliers",
Proceedings of the 2000 ACM SIGMOD international conference on Management of
data, pp. 93-104, https://doi.org/10.1145/342009.335388
[38] Liu, F. T., et al., 2008, "Isolation forest", 2008 eighth ieee international conference
on data mining, pp. 413-422, IEEE. https://doi.org/10.1109/ICDM.2008.17
[39] Hariri, S., et al., 2019, "Extended isolation forest", IEEE transactions on
knowledge and data engineering,Vol. 33(4), pp. 1479-1489.
https://doi.org/10.1109/TKDE.2019.2947676
[40] Motion Control Business Unit, I. D. B. D., Panasonic Industry Co., Ltd., 2022,
Technical Reference - Functional Specification - Panasonic Industry Co., Ltd. .
https://mediap.industry.panasonic.eu/assets/download-
files/import/mn_minas_a6b_ethercat_functional_specification_pidsx_en.pdf
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top