臺灣博碩士論文加值系統

無人車最初主要應用於物流和製造業，然而現今已廣泛擴展至倉儲、醫療、食品行業等多個領域。在早期的無人車應用中，常須依賴磁條來規畫移動路徑。通常情況下，當工廠內設備已完成定位後，才會開始鋪設磁條以供無人車使用，然而，這種方法的鋪設時間隨著工廠大小的增加而增加。此外，由於長時間使用後，磁條的磁性可能會逐漸衰減，導致無人車難以感應到磁條，進而影響無人車的正常運作。為了克服這些問題，引入虛實整合技術作為一種解決方案。虛實整合是一種虛擬模型，用於模擬現實世界中的物體、系統或過程。透過虛實整合，可以先建立工廠的虛擬模型，並在虛擬環境中規劃與測試無人車的移動路徑。因此本研究採用虛實整合系統輔助無人車的路徑規劃，用於解決早期無人車受限於技術和應用場域，需要依靠磁條規畫路徑和無法提前規劃路徑等問題。首先在虛實整合系統中建立虛擬環境，包括工廠、加工機和無人車利用物體重疊檢測、虛擬模型約束力和無人車運動學模型等技術，實現虛擬模型碰撞、運動再現和控制功能，用於模擬和規劃無人車的移動路徑。透過TCP/IP通訊協議和SLMP通訊協議建立實體與虛擬雙向數據交換系統，記錄實體無人車的馬達轉數以獲得無人車的數位模型，用於控制虛擬無人車的移動，或者根據虛擬無人車的移動數據，透過該系統反向控制實體無人車的移動。使用python在虛實整合系統中建立控制介面，用於模式切換、顯示數據、手動控制和規劃路徑等功能。藉由在虛擬環境反覆測試無人車以不同的切入角規畫路徑，並移動至工具機前方，獲得應使用何種數值，以達到最接近於工具機前方的目的，經實驗得知使用150度的切入角，與工具機距離為0.037公尺。接著根據實驗數值和無人車當前路徑起點，規劃六種路徑用於應對無人車於該虛擬環境中的移動。最後使用虛實整合系統遠端控制虛實整合無人車，分析規畫路徑中直線和圓弧路徑，應使用何種速度值，以達到最小定位誤差，經實驗得知直線移動時使用0.1m/s移動，對於網路延遲造成的影響最小，但移動時間是全部移動速度中最長，當移動距離為25公尺時，移動誤差為0.071公尺，移動時間為250秒，因此對於須移動5公尺以上直線距離時，應先使用1m/s速度移動，直到剩餘距離為1公尺時，在切換為0.1m/s移動，藉由在一段路程中混合兩種速度已達到移動時間短且移動誤差最小。

Unmanned vehicles were initially primarily used in logistics and manufacturing industries but have now expanded to various fields such as warehousing, healthcare, and the food industry. In the early applications of unmanned vehicles, they often relied on magnetic strips to plan their movement paths. Typically, after the positioning of equipment within a factory was completed, magnetic strips were laid for the unmanned vehicles to use. However, the time required for this strip laying process increased with the size of the factory. Additionally, over prolonged usage, the magnetic properties of the strips might gradually decay, making it difficult for the unmanned vehicles to detect them, thereby affecting their normal operation.To overcome these issues, the integration of virtual and real (VR) technologies is introduced as a solution. Virtual-Reality (VR) integration is a technique that utilizes a virtual model to simulate objects, systems, or processes in the real world. Through VR integration, a virtual model of the factory can be created to simulate and test the movement paths of unmanned vehicles in a virtual environment. In this study, a VR integration system is used to assist in the path planning of unmanned vehicles, addressing issues such as early unmanned vehicles being constrained by technology, relying on magnetic strips for path planning, and the inability to plan paths in advance.Firstly, a virtual environment is established in the VR integration system, including the factory, machining equipment, and unmanned vehicles. Techniques such as object overlap detection, virtual model constraint forces, and the kinematic model of unmanned vehicles are employed to achieve collision simulation, motion reproduction, and control functions in the virtual model for simulating and planning the movement paths of unmanned vehicles. A bidirectional data exchange system between the real and virtual entities is established using TCP/IP communication protocols and SLMP communication protocols. This system records the motor speeds of physical unmanned vehicles to obtain digital models of the unmanned vehicles, which are then used to control the movement of virtual unmanned vehicles. The system can also control the physical unmanned vehicles based on the movement data of virtual unmanned vehicles. A control interface is created in the VR integration system using Python for functions such as mode switching, data display, manual control, and path planning.Through iterative testing in the virtual environment, unmanned vehicles are planned with different approach angles to simulate and move to the front of the machine tool, obtaining the values needed to achieve the closest proximity to the machine tool. Through experiments, it is found that using a 150-degree approach angle with a distance to the machine tool of 0.037 meters achieves the desired goal. Subsequently, six paths are planned based on the experimental values and the current starting point of the unmanned vehicle to address the movement in the virtual environment. Finally, the VR integration system is used to remotely control the unmanned vehicle in the VR integration system, analyzing the planned paths with straight-line and circular trajectories. The appropriate velocity values are determined to minimize positioning errors. Through experimentation, it is found that a speed of 0.1 m/s is suitable for straight-line movement, minimizing the impact of network delays. However, this speed results in the longest travel time among all speeds. For distances greater than 5 meters, it is advisable to initially use a speed of 1 m/s until the remaining distance is 1 meter, and then switch to 0.1 m/s movement. This hybrid approach minimizes both travel time and positioning errors for straight-line distances.

摘 要.................i
Abstract.................ii
誌 謝.................iv
目 錄.................v
表 目 錄.................vii
圖 目 錄.................viii
符號說明.................x
第 一 章 緒 論.................1
1.1 研究背景與動機.................1
1.2 研究目的.................1
1.3 論文架構.................1
第 二 章 文獻回顧.................3
2.1 虛實整合應用於工具機之相關研究.................3
第 三 章 研究架構與方法.................10
3.1 研究架構與流程.................10
3.2 建置與組裝設備與整場虛實整合平台.................11
3.2.1 工廠3D模型演示.................11
3.2.2 工廠和設備的3D建模與組合.................12
3.3 視覺渲染擬真優化與定義材料特性.................15
3.3.1 虛實整合機器人模擬軟體.................15
3.3.2 工廠與設備模型優化.................17
3.4 3D虛擬物理引擎及追蹤模擬計算.................21
3.4.1 物體重疊檢測原理.................21
3.4.2 建立虛擬模型運動方式.................26
3.4.3 無人車運動方程式推導.................31
第 四 章 實驗結果與討論.................33
4.1 虛實整合系統通訊與控制.................33
4.1.1 雙向數據交換系統建置.................33
4.1.2 網路通訊架構.................34
4.2 遠端控制介面與虛擬路徑規劃.................37
4.2.1 遠端控制介面.................37
4.2.2 路徑規劃.................38
4.3 無人車虛實整合遠端路徑定位控制.................41
第 五 章 結論與未來展望.................45
5.1 結論.................45
5.2 未來展望.................45
參考文獻.................46
Extended Abstract.................48

[1] Cozmiuc, D., Petrisor, I., 2018, “Industrie 4.0 by siemens: Steps made today”, Journal of Cases on Information Technology, Volume 20, Issue 2, April, Pages 30-48.
[2] Alcácer, V., Cruz-Machado, V., 2019, “Scanning the Industry 4.0: A Literature Review on Technologies for Manufacturing Systems”, Engineering Science and Technology, an International Journal, Volume 22, Issue 3, June.
[3] Moghaddam, M., et al., 2018, “Reference architectures for smart manufacturing: A critical review”, Journal of Manufacturing Systems, Volume 49, October, Pages 215-225.
[4] Yun, J.P., et al., 2020, “Automated defect inspection system for metal surfaces based on deep learning and data augmentation”, Journal of Manufacturing Systems, Volume 55, April, Pages 317-324.
[5] Ren, L., et al., 2018, “Bearing remaining useful life prediction based on deep autoencoder and deep neural networks”, Journal of Manufacturing Systems, Volume 48, July, Pages 71-77.
[6] Fuller, A., et al., 2020, “Digital Twin: Enabling Technologies, Challenges and Open Research”, IEEE Access, Volume 8, Pages 108952-108971.
[7] Fei T., et al., 2019, “Digital Twin in Industry: State-of-the-Art”, IEEE Transactions on Industrial Informatics, vol. 15, no. 4, April, pp. 2405-2415.
[8] Jay L., Behrad B., Hung-An K., 2015, “A Cyber-Physical Systems architecture for Industry 4.0-based manufacturing systems”, Manufacturing Letters, Volume 3, January, Pages 18-23.
[9] Abid I., et al., 2022, “Modeling, Analysis, and Optimization of Robotic Light Machining Tasks for Empowering Digital Twin: Generalized Impulse Model Approach”, IEEE Access, Vol. 10, pp. 105133-105148.
[10]Juntong Y., et al., 2023, “Digital twin model construction of robot and multi-object under stacking environment for grasping planning”, Volume 149, Part A, December
[11]Mingyang L., et al., 2023, “Research on the Construction of a Digital Twin Model for a 3D Measurement System of Gears”, IEEE Access, Vol. 11., pp. 124556-124569.
[12]Rongli, Z., et al., 2019, “Digital Twin-Driven Cyber-Physical System for Autonomously Controlling of Micro Punching System”, IEEE Access, vol. 7, pp. 9459-9469.
[13]Zhaoqian W., et al., 2023, “Digital Twin-Driven 3-D Position Information Mutuality and Positioning Error Compensation for Robotic Arm”, IEEE Sensors Joumal, Volume 23, no. 22, November 15, pp. 27508-27516.
[14]Nursultan J., et al., 2023, “Integrating Machine Learning Model and Digital Twin System for Additive Manufacturing”, IEEE Access, vol. 11, pp. 71113-71126.
[15]Kuo L., et al., 2022, “Time-Varying Error Prediction and Compensation for Movement Axis of CNC Machine Tool Based on Digital Twin”, IEEE Transactions on Industrial Informatics, vol. 18, no. 1, January, pp. 109-118.
[16]Lilan, L., et al., 2022, “Digital twin-driven surface roughness prediction and process parameter adaptive optimization”, Advanced Engineering Infornatics, Volume 51, January.
[17]Tao F., et al., 2019, “Five-dimension digital twin model and its ten applications”, Comput. Integr. Manuf. Syst., vol. 25, no. 1, January, pp. 1-18.
[18]Zhen-yuan J., et al.. 2018, “A review of contouring-error reduction method in multi-axis CNC machining”, International Journal of Machine Tools and Manufacture, Volume 125, 20 August, Pages 34-54.
[19]Andrew Witkin, 1997, “Physically Based Modeling: Principles and Practice Constrained Dynamics”.
[20]Michael G., 2002, “SME Management Forum Completing the Cycle: Using PLM Information in the Sales and Service Functions”.
[21]Andrea de Giorgio, et al., 2017, “Human-machine collaboration in virtual reality for adaptiveproduction engineering”, 27th International Conference on Flexible Automation and Intelligent Manufacturing, June.
[22]Yuxi, Z.,et al., 2022, “Digital Twins Visualization of Large Electromechanical Equipment”, IEEE Journal of Radio Frequency Identification, vol. 6, pp. 993-997.