臺灣博碩士論文加值系統

隨著工具機產業走入工業4.0與物聯網的時代，其中智慧工具機投入智慧製造目的是為了提升生產效率與加工品質，製造業對於工具機的加工精度要求也越來越高，許多廠家為了增加生產效率與產量使工具機需拉長加工時間，如此長時間的工作必會讓工具機部件產生熱源，故本研究基於類神經演算法之疊代收斂性預測與補償螺桿熱變形研究以不同溫度變化下準確地補償誤差值減少熱誤差為主要目的，研究論文主要分成四個部分，首先須在工具機上建置溫度量測模組，包含埋設溫度感測器來量測長時間運轉後進給螺桿之溫度變化，搭配雷射干涉儀檢驗螺距誤差且分析溫度與螺桿熱變形量之間的關係，第二部分是將溫升之螺桿熱變形與行程作為輸入資料利用曲面擬合與倒傳遞神經網路建立預測模型，其中曲面擬合使用Gauss與Logistic演算法進行擬合，倒傳遞神經網路使用最陡坡降法、Adagrad、RMSProp、Momentum與Adam之疊代方式分析其演算過程對誤差函數的影響，第三部分是藉由自行開發之遠端智能化補償系統將螺桿熱變形誤差進行相對應的補正達到減少熱誤差之目的，其中由預測誤差值經過補正值公式計算補正值，透過遠端監控技術將補償值輸入至控制器完成補償，最後一部分是建立資料庫與雲端管理系統，就由此系統讓管理者方便蒐集機台溫度變化與操作資訊，將量測後溫度值及重要加工資料存入資料庫，透過雲端看板可以隨時地監視機台狀態，整合檢驗、機器學習、補償以及物聯網技術完成工具機智能化之能力來提升機台加工品質。

The CNC machine tool industry enters the era of industrial 4.0 and Internet of Things(IOT). The smart manufacturing improves production efficiency and processing quality. The Precision of the CNC machine tool has higher requirments for the manufacturing. Many manufacturers want to increase efficiency and output of production. Then the CNC machine tool worked after long time. The temperature of the screw has increased. That create thermal deformation for the screw. The thermal error is important for precision of the CNC machine tool. The method solves the thermal deformation by cooling system and compensation system. Although using cooling system reduce most thermal deformation. But that needs more cost. Therefore, the search used machine learning for prediction thermal deformation of the screw. And the thermal deformation has compensated by intelligent compensation system at different temperature of the screw. The search divided into four parts. Firstly, using thermocouple sensor established measurement module at the feed system of CNC machine tool. This measurement module measured the temperature of the screw bearing. Then the thermal deformation has inspected different temperature. Secondly, the temperature and travel of the axis input to models for prediction. The prediction model of surface fitting (SF) and backpropagation neural network (BPN) predicted the thermal deformation of the screw. SF include the algorithms of Gauss and Logistic. BPN is analyzed the influences for error function by algorithms of steepest gradient, Adagrad, RMSProp, Momentum, and Adam. Thirdly, the method of compensation used the self-developed remote intelligent compensation system for reducing the thermal deformation of the screw. The correction formula calculates the value of correction for the pitch error. The value of correction inputed to the controller. Finally, using software, Postgre SQL and Node-red, builds the database and cloud monitor. The user can be convenient to manage the information of CNC machine tool with the database and cloud monitor. The information of the machine includes the temperature of the screw bearing, feed rate, program name, and cycle time, etc. The CNC machine tool becomes an intelligent machine with inspected, machine learning, prediction, compensation, and IOT for upgrading processing quality.

摘要...i
Abstract...ii
誌謝...iii
目錄...iv
表目錄...vi
圖目錄...vii
符號說明...ix
第 一 章 緒論...1
1.1 研究背景與動機...1
1.2 研究目的...2
1.3 論文架構...2
第 二 章 文獻回顧...3
2.1 工具機進給螺桿熱變形與補償相關研究...3
2.2 機器學習預測演算法之應用相關研究...7
第 三 章 研究架構與方法...13
3.1 研究架構與流程...13
3.2 硬體及系統架構...14
3.2.1 準力JL-20305X五軸加工機...14
3.2.2 Renishaw XL-80雷射干涉儀...14
3.2.3 溫度量測模組...15
3.3 工具機誤差介紹...16
3.3.1 線性軸誤差...16
3.3.2 背隙誤差...17
3.4 溫度影響螺桿熱變形檢驗與分析...18
3.4.1 硬體架設...18
3.4.2 雷射干涉儀架設...18
3.4.3 雷射干涉儀原理...19
3.4.4 ISO檢驗標準與補償值計算...20
3.5 曲面擬合預測模型...22
3.6 機器學習之倒傳遞神經網路...23
3.6.1 資料前處理...23
3.6.2 倒傳遞神經網路...24
3.6.3 疊代演算法...28
3.7 遠端監控與智能化補償系統...32
3.7.1 智研控制器通訊...32
3.7.2 智能化補償規則...34
3.8 雲端資料庫管理開發...35
第 四 章 實驗結果...38
4.1 溫度影響螺桿熱變形檢驗與分析...38
4.1.1 工具機進給系統精度檢驗...38
4.1.2 溫度變化量測結果...40
4.1.3 螺桿熱變形檢驗結果探討...40
4.2 曲面擬合預測螺桿熱變形...41
4.3 倒傳遞類神經網路學習與預測...44
4.3.1 倒傳遞神經網路之疊代演算法學習效果分析...44
4.3.2 倒傳遞神經網路不同疊代演算法之預測結果...52
4.4 智能化補償螺桿熱變形...54
4.4.1 X軸螺桿熱變形補償...55
4.4.2 Y軸螺桿熱變形補償...56
4.4.3 Z軸螺桿熱變形補償...56
4.5 遠端監控與智能化補償系統介面...57
4.5.1 遠端監控系統介面...57
4.5.2 智能化補償系統介面...58
4.6 雲端資料庫管理系統...59
4.6.1 雲端看板...59
4.6.2 資料庫管理系統...60
第 五 章 結論與未來展望...62
參考文獻...63
附件一...65
Extended Abstract...66

[1]Intel, ”各類產業的物聯網應用”
[2]財團法人精密機械發展中心,”工具機熱誤差與因應對策”,2016
[3]每日頭條,”了解人工智慧之分類和預測-數據中的預測、分類和迴歸分析”, 2017
[4]Hu Shi,” Investigation into effect of thermal expansion on thermally induced error of ball screw feed drive system of precision machine tools”, International Journal of Machine Tools & Manufacture, pp. 60-71, 2015
[5]Wenlong Feng,” Thermally induced positioning error modelling and compensation based on thermal characteristic analysis”, International Journal of Machine Tools & Manufacture, Volume 93, Pages26-36, June 2015
[6]Chao Jin,” Thermal characteristics of a CNC feed system under varying operating conditions”, Precision Engineering, Volume 42, October 2015, Pages 151-164
[7]黃昌隆 ,”單軸移動平台定位精度補償方法之研究”, 國立台北科技大學, 碩士學位論文 ,2016
[8]Massimiliano Patacchiola,” Head pose estimation in the wild using Convolutional Neural Networks and adaptive gradient methods”, Pattern Recognition, Volume 71, November 2017, Pages 132-143
[9]A. Shebani,” Prediction of wheel and rail wear under different contact conditions using artificial neural networks”, Wear, Volumes 406-407, July 2018, Pages 173-184
[10]Mozammel Mia,” Prediction of surface roughness in hard turning under high pressure coolant using Artificial Neural Network”, Measurement, Volume 92, October 2016, Pages 464-474
[11]Alvaro Arcos-Garcia,” Deep neural network for traffic sign recognition systems: An analysis of spatial transformers and stochastic optimisation methods”, Neural Networks, Volume 99, March 2018, Pages 158-165
[12]林伯尚 ,”基於倒傳遞類神經網路之刀具壽命與表面粗糙度研究” ,國立虎尾科技大學 ,碩士學位論文 ,2018
[13]Amin Najafi,” Performance prediction of hydrofoil- supported catamarans using experiment and ANNs”, Applied Ocean Research, Volume 75, June 2018, Pages 66-84
[14]Z.Z. Xu,” Thermal error forecast and performance evaluation for an air-cooling ball screw system”, International Journal of Machine Tools & Manufacture, Volume 51, Issues 7-8, pp. 605-611, July-August 2011
[15]Cheng-Hsien Wu,” Thermal analysis for the feed drive system of a CNC machine center”, International Journal of Machine Tools & Manufacture, Volume 43, Issues 15, pp. 1521-1528, December 2003
[16]吳威志,”C-Type機型之五軸同動誤差檢測分析與補償”, 國立虎尾科技大學, 碩士學位論文, 2017
[17]英國Renishaw公司,台灣官網XL-80雷射系統
[18]炘珈儀表官網，” https://www.sj-gauge.com/_tw/faq/detail.php?cid=15&faqid=167”
[19]財團法人精密機械發展中心,”五軸工具機誤差量測技術”
[20]ISO 230-3, Acceptance code for machine tools part 3:
[21]ISO 230-2, Acceptance code for machine tools part 2: determination of accuracy and repeatability of positioning of numerically controlled machine tools, International Standard,1988.
[22]蘇春維,”工具機熱變位補償應用技術”, 財團法人精密機械發展中心,2016.
[23]葉怡成，類神經網路模式應用與實作，儒林圖書有限公司，1993 年 1 月初版，2000 年 4 月七版.
[24]Tommy Huang,”機器/深度學習-基礎數學(三):梯度最佳解相關算法”, 2018
[25]Diederik P. Kingma,” ADAM: A METHOD FOR STOCHASTIC OPTIMIZATION”, 2015
[26]智研科技有限公司網站 ,技術支援IClient編程說明
[27]Cubie, MQTT教學(一):MQTT, 2017
[28]何信昱, IOT視覺化開發工具, MAKERPRO , 2017
[29]N.E. Karkalos,” Surface roughness prediction for the milling of Ti–6Al–4V ELI alloy with the use of statistical and soft computing techniques”, Measurement, Volume 90, August 2016, Pages 25-35
[30]馬政剛,“類神經網路誤差估計與學習速率之探討”, 逢甲大學-應用數學系碩士班, 2013
[31]Wen-Tsao Pan,” The Study On The Optimizing Predictability Of The Neural Networks: Case Study Of Finance Holding Stocks”, 遠東學報第二十四卷.
[32]Han-Xiong Huang, Jiong-Cheng Li, Cheng-Long Xiao,” A proposed iteration optimization approach integrating backpropagation neural network with genetic algorithm”，Procedia Computer Science, Volume 45, Volume 42, Issue 1, January 2015, Pages 146-155.
[33]林俊良, “智慧型控制分析與設計”,全華圖書股份有限公司,第3版