臺灣博碩士論文加值系統

製造業近年來對於加工機故障或刀具磨耗之預測模型開發日趨重視。過往數據分析的刀具磨耗監測系統研究中，大部分是根據已經制定好的切削條件(轉速、進給)去建立刀具磨耗監測模型的，使用此模型對監測模型對應的切削條件之量測數據有良好的辨識性。但對於此切削條件外的數據的辨識性能力之研究不多，因此此方法建立的模型之泛化性(generalization)仍需要驗證。本研究嘗試將特定切削條件下的數據排除於訓練數據集中，並作為測試樣本，測試模型之辨識率，藉此檢驗刀具磨耗監測模型之泛化性。
研究中利用置於虎鉗上的加速規擷取之振動訊號，另外以電流勾錶量測主軸電流訊號，並將這些訊號作為刀具磨耗狀態的對應特徵。加速規安裝空間需求小、安裝容易、成本低，且有較高的信噪比(Signal-to-Noise ratio)，相較於安裝動力計更適合用於量測小型五軸加工機之加工訊號，而主軸電流訊號與切削力有高度正相關。本研究利用上述特徵以隨機森林(Random forest)演算法建立機器學習模型來辨識刀具磨耗之狀態，為測試模型之泛化性，採用排除特定實驗條件於訓練集外而作為測試集的方式測試模型，對各種實驗條進行測試後準確率大部分超越80%，最低則是65%。對實驗數據事先進行標準化後，去除頻率域訊號能量的因素，使模型選取到結構振動特徵，可將最低的準確率提升至75%，另外也解決了將嚴重磨耗判斷為輕度磨耗的問題，可以正確地判斷換刀時機。

In recent years, the manufacturing industry has paid more and more attention to the development of models for predicting machine failure or tool wear. In the past, there are researches studying tool wear monitoring system based on data analysis. Most of them built tool wear monitoring model using cutting conditions, for example speed and feed, according to ones they had already set. By using these models, we can correctly identify the testing data which correspond to the condition they set. In contrast, few researches that studying identifying the data which is not correspond to the condition we set. Therefore, generalization of models using these methods still needs to be discussed. In this research, data corresponding to specific cutting condition is excluded from training data set. Then, these data will be included to testing data set to test the accuracy of the model. The generalization of tool wear monitoring model could be test by this method.
The accelerometer is attached on vise to capture vibration signal, and the current clamp meter is used to capture spindle current signal. These signal will be the features of their tool wear condition. Accelerometer is small, easy to install, relatively low cost and is has high signal-to-noise ratio. Using vibration signal can monitor the change of tool status more instantaneously than the cutting force. The spindle current has highly positive correlation with cutting force. In this study, vibration and current signal will be used to build a random forest model to identify the wear status. In order to test the generalization of the model, some data capture in specific cutting condition is exclude from training set but include in testing set. The result is that the accuracies are higher than 80% in almost all conditions, and the lowest on is about 65%. After standardization, the power factor will be removed so that features of structure vibration when feature selecting, and the lowest accuracy can increase to 75%, in addition, the problem that model fail identified serious wear into little wear is solved. Therefore, the model can identify tool wear status more correctly.

口試委員會審定書 I
誌謝 II
摘要 III
ABSTRACT IV
目錄 VI
圖目錄 IX
表目錄 XII
第1章 緒論 1
1.1研究背景 1
1.2研究動機與目的 2
1.3研究架構 4
第2章 文獻回顧 5
2.1 切削過程之訊號監控 5
2.2刀具磨耗監控 7
2.3小結 8
第3章 研究方法 9
3.1研究架構 9
3.2 訊號擷取與分析 11
3.2.1 感測器之裝設 11
3.2.2 奈奎斯特定理(Nyquist Theorem) 11
3.2.3 訊號前處理 13
3.2.4 訊號頻域轉換 16
3.2.5 功率譜轉換(Welch Method) 18
3.3隨機森林(RANDOM FOREST) 19
3.3.1隨機森林概念 19
3.3.2 決策樹 19
3.3.3 隨機森林模型架構 21
3.4特徵工程 22
3.4.1 特徵集合 23
3.4.2 特徵標準化 25
3.4.3 隨機森林特徵選取 27
3.5刀具磨耗之分析 29
3.5.1 刀具磨耗量測 29
3.5.2 磨耗等級標籤 31
第4章 實驗設備與規劃 33
4.1實驗架構 33
4.2實驗設備 34
4.3實驗規劃 43
4.3.1銑削實驗流程 43
4.3.2訊號分析流程 44
第5章 實驗結果與討論 46
5.1隨機森林分類結果 46
5.1.1模型準確率測試 47
5.1.2 分類結果分析 49
5.1.3數據量多寡之影響 56
5.1.4 特徵工程改善 62
5.1.5 小結 70
5.2重要特徵分析 72
5.2.1 重要特徵比較 72
5.2.2 標準化對重要特徵之影響 74
5.2.3小結 77
第6章 結論與未來展望 78
6.1總結 78
6.2未來展望 79
參考文獻 80

[1] ElMaraghy, Hoda, “Evolution and future of manufacturing systems,” CIRP Annals - Manufacturing Technology, pp. 1-24, 20 july 2021.
[2] L. Monostori, and H. Van Brussel, “Machine Learning Approaches to Manufacturing,” CIRP Annals, pp. 675-712, 1996.
[3] M.C.Shaw, Metal cutting prsinciple, New York: Oxford University Press, 2005.
[4] N. Ambhore, D. Kamble, S. Chinchanikar, and V. Wayal, “Tool condition monitoring system: A review,” Materials Today: Proceedings, pp. 3419-3428, 2015.
[5] M. Malekian, S. S. Park , and M. B. G. Jun, “Tool wear monitoring of micro-milling operations,” Journal of Materials Processing Technology, pp. 4903-4914, 2009.
[6] J. T. Roth, and S. M. Pandit, “Monitoring End-Mill Wear and Predicting Tool Failure Using Accelerometers,” Journal of Manufacturing Science and Engineering, pp. 559-567, 1999.
[7] J. Z. Zhang, and J. C. Chen, “Tool condition monitoring in an end-milling operation based on the vibration signal collected through a microcontroller-based data acquisition system,” International Journal of Advanced Manufacturing Technology, pp. 118-128, 200.
[8] Rodolfo E. Haber, Jose E. Jiménez, C. Ronei Peres, and José R. Alique, “An investigation of tool-wear monitoring in a high-speed machining process,” Sensors and Actuators, pp. 539-545, 2004.
[9] Milla Caroline Gomes, Lucas Costa Brito, and Marcio Bacci da Silva, “Tool wear monitoring in micromilling using Support Vector Machine with vibration and sound sensor,” Precision Engineering, pp. 137-151, 2021.
[10] P. Krishnakumar, K. Rameshkumar, and K. I. Ramachandran, “Tool Wear Condition Prediction Using Vibration Signals in High Speed Mechining(HSM) of Titanium(Ti-6Al-4V) Alloy,” Procedia Computer Science, pp. 270-275, 2015.
[11] A D Patange, Jegadeeshwaran R, and N C Dhobale, “Milling cutter condition monitoring using machine learning approach,” IOP Conf. Series: Materials Science and Engineering, 2019.
[12] 王崇穎, and 盧銘詮, 回饋型隱藏式馬可夫模型於鎳基材料切削刀具磨耗多重感測器偵測系統開發, 台中: 國立中興大學, 2018.
[13] 蔡乙陞, and 李貫銘, 銑削加工振動訊號應用於刀具磨耗監控之研究, 台北: 國立台灣大學, 2019.
[14] C. L. Phillips, J. M. Parr, and E. A. Riskin, Signals, systems, and transforms, 3rd ed, Upper Saddle River, N.J: Prentice Hall, 2003.
[15] By Tukey, James W. Cooley and John W., “An Algorithm for the Machine Calculation of Complex Fourier Series”.
[16] L. Breiman, “Random Forests,” Machine Learning, pp. 5-32, 2001.