RTAI+: 針對CNC程序特性的RTAI即時排程修補機制

Abstract

工具機是當前製造產業的重點工具之一，主要應用於自動化流程來達到高效率生產。工業4.0 提出的概念更結合了智慧感控系統與物聯網等新的電腦網路技術於自動化控制產業。相較傳統的自動化設備，工業4.0 更注重即時運算能力與網際網路的連線效能。也因此開啟了PC-based 控制器的發展空間。在本論文中，我們使用IgH EtherCAT Master 建立了以EtherCAT為通訊基礎的PC-based 控制器，並安裝即時性套件RTAI以增進PC系統的即時性能力。為了進一步增進控制系統的穩定性，我們針對工具機控制系統的基本特性進行研究，並找出CNC 程序的特定工作圖樣。我們依此發現設計了重排工作圖樣(rearranged pattern) 的機制，並提出RTAI+，一種以RTAI 為基礎，專用於CNC程序的即時性套件。我們將RTAI+ 應用於IgH EtherCAT Master，並調整部分模組提出IgHXRTAI+ 改善CNC 程序的即時性與穩定性。為了驗證整體系統，我們設計多種負載測試與多程序測試實驗。在兩種實驗中，使用RTAI+系統的最大時間抖動量都可以控制在3% 以內(<30s) 且其時間抖動量標準差皆小於使用原RTAI 的系統，顯示RTAI+ 提高的系統穩定性。另外，在低負載的情況下，搭配RTAI+ 的IgH EtherCAT Master也能進一步地降低時間抖動量的標準差，使系統的即時性更加穩定。從多程序測試中，我們則驗證RTAI+ 運行時，系統中的各個程序的穩定性是一致的，並不會隨著執行順序改變，這也驗證了RTAI+ 可適用於同時監控多個工具機的情境。

Machine tool is one of the major technologies in manufacturing industries. The highprecise machine tools are used in automated production and provide better efficiency in manufacturing. In recent years, Industry 4.0 is presented by integrating devices for real-time computation and for Internet connection. This makes PC-based controllers more potential in the future. In this paper, we adapt IgH EtherCAT Master to create a PC-based CNC controller supporting EtherCAT. The system also patches RTAI to enhance the real-time capability. To improve the stability of our real-time system, we propose RTAI+ and a new pattern (rearranged pattern) for motor tasks. We also apply RTAI+ to IgH EtherCAT Master as IgH RTAI+ to enhance the stability of transmission processes. To verify the stability of our system, we design several loading tests and multi-task tests. In the two kinds of experiments, the maximum time jitter of the system using RTAI+ is smaller than 3% (<30s). And in the system using RTAI+, the standard deviation of jitter is smaller than that using RTAI. This confirms that RTAI+ can improve the stability of a PC-based CNC system. On the other hand, When applying RTAI+ to IgH EtherCAT Master, the standard deviations can be further reduced in the low loading situations. The multi-task test also shows that the RTAI+ system has nearly the same standard deviations in each processes. In te multi-task tests, we prove that most processes running on the system using RTAI+ have the nearly same deviations. This proves that the proposed RTAI+ can provide stability in controlling multiple machine tools simultaneously.