爪哇手機控制遠端製造系統的微員族架構設計

Abstract

本論文利用微員族概念建立出監控員、偵查員、動畫員、資料伺服員、虛擬製造系統和模擬器的架構被提出來解決以Java手機遠端控制製造系統的問題。在建構一套監控與偵察製造生產系統的軟體時，通常會參照五項步驟來實作。在監控的層面來說，有四項步驟用來規劃如何定義詳細軟體的需求規格、如何驗證軟體的設計可行性、如何撰寫同步併發式平行架構的應用程式，以及為控制製造系統正常運作之時，該如何測試軟體相互運行的狀態。在偵察的層面，主要任務是怎麼設計出能立刻查出個別出現的反常事件的軟體。在本論文，一支具備可程式的Java語言手機被使用來監控與偵察一套遠端的製造系統。從這個觀點，這五項步驟必須進一步的採行附加的無線考量，因此也導致產生額外的程式應用工作，譬如J2ME的程式為了人手接聽電話介面與無線通訊而設計，J2SE程式在可程式邏輯控制器(PLC)、系統模擬器中編寫程式等。經由研究結果提出並分別設計了監控員、偵查員、動畫員、資料伺服員、虛擬製造系統以及系統模擬器六個軟體模組應用來解決上述的控制遠端的製造系統問題。在所持用手機與指定製造系統的成功整合下，顯示出針對微員族架構提出之合理性。

A homunculus-oriented framework consisting of supervisor, monitor, animator, data server, virtual manufacturing system, and simulator are proposed to solve the remote manufacturing system control problem through a java-based mobile phone. Traditionally five tasks are involved to implement software for supervising and monitoring a manufacturing system. At the supervising phase, the four tasks are how to specify the software requirements, how to validate the software design, how to code concurrent programs, and how to test the interactive software behaviors for controlling the normal manufacturing activities. At the monitoring phase, the main task is how to design the software for detecting an emerging abnormal event immediately. In the thesis, a Java-based mobile phone is used to supervise and monitor a remote manufacturing system. From this viewpoint, these five tasks have to be progressed with additional wireless considerations. They also require extra software efforts such as J2ME programming for man-phone interface design, wireless communications, J2SE programming within Programming Logic Controller (PLC), manufacturing system simulation, etc. As a research result, six software modules named supervisor, monitor, animator, data server, virtual manufacturing system, and simulator are respectively proposed, designed, and implemented to solve the remote control problems above. The successful integration of the given mobile phone and the underlying manufacturing system shows the feasibility of this six- homunculus framework approach.