智慧噴塗路徑生成技術之研究

智慧噴塗技術是現代塗裝產線智慧升級的重要方向，傳統噴塗路徑設計仰賴人工經驗，缺乏靈活性與一致性，易造成粉末浪費與品質不穩定。本文針對異形工件之智慧噴塗需求，開發整合數位雙生模擬、AI影像辨識與智慧噴塗路徑自動生成之系統，實現噴塗流程智慧化。透過Visual Components建立虛實對應產線模型，配合GAN生成高品質虛擬影像訓練YOLOv8辨識模型，快速獲取工件特徵，並以3D光學掃描提升噴塗軌跡精準度。智慧路徑生成演算法動態控制4支噴槍，搭配即時監控靜電值、噴塗量與粉末回收系統，達成高效率、低能耗之智慧噴塗效果。實務驗證顯示，系統可提升生產效率32%，粉末利用率達80%以上，良率穩定98%，有效促進塗裝產線節能減碳與智慧升級。

In modern coating production lines, intelligent spray technology is a key direction for smart upgrading. Traditional spray path design relies on manual experience, lacking flexibility and consistency, which easily causes powder waste and unstable quality. This study focuses on the intelligent spray requirements of complex-shaped workpieces, developing an integrated system combining digital twin simulation, AI image recognition, and automatic spray path generation. A virtual-real production line model was built using Visual Components. GAN was used to generate high-quality virtual images to train YOLOv8 recognition models, enabling rapid acquisition of workpiece features. Combined with 3D optical scanning, the precision of spray trajectories was significantly improved. The intelligent path generation algorithm dynamically controls four spray guns with real-time monitoring of electrostatic values, spray volume, and powder recovery system, achieving highly efficient and low-energy-consumption intelligent spraying. Practical verification shows that the system improves production efficiency by 32%, powder utilization rate exceeds 80%, and yield remains stable at 98%, effectively promoting energy-saving, carbon-reducing, and intelligent upgrading of coating production lines.