基於邊緣安全監測的工業物聯網應用 - 自動引導搬運車運作的案例研究

Abstract

隨著工業物聯網IIoT（Industrial IoT）的發展，工廠的自主系統變得更加開放、動態、彈性、適應性及複雜化。這些系統中，最基本安全的要求應該意識到環境的變化，對於會產生負面影響的變化，在實施反饋閉環中不斷地調整系統的行為。為了確保自主系統在IIoT環境中的安全性，應將安全驗證移至運行時。在本論文中，為了處理安全運行時監控，我們通過使用Event-B的描述和反應代理模型，提出了安全規則的模型驅動方法。該方法將監控系統的安全規範轉換到在Event-B語法的描述中，透過使用Event-B的設計，進而將中間轉換過程的安全規範轉換成安全監控代理的規則。另外，我們還提出了監控架構，並使用參考AGV工業安全標準的案例來測試監控系統的表現。

With the development of IIoT (Industrial IoT), the autonomous system of the factory becomes more open, dynamic, flexible, adaptable and complex. The fundamental safety requirement of those systems should be aware of the changes in the environment that affect negatively and be implemented the feedback closed loop that continuously adjust the behavior of systems. To ensure safety of the autonomous system in IIoT environment, it is desirable to shift safety verification to runtime. In this thesis, to deal with the safety runtime monitoring, we proposed the model-driven approach of the safety rules by using the Event-B specification and reactive agent model. This approach transforms from specifying the safety rule for the monitored system to the design method in Event-B specification, using the design of Event-B to support the intermediate transformation process from the safety rule to the rule of the safety monitoring agent. We also proposed our monitoring architecture and use a case study in reference to AGV industrial safety standard to test the monitoring system.