摘要

對於輸送機而言，最受到關注且嚴重的災害為火災事件，因此如何發展一套方法，來有效地達到提前檢測並預警的功能，一直以來是輸送機使用者的需求，亦是許多研究者投入發展的策略與方向。然而，輸送機火警發生的原因相當多元，僅仰賴單一感測技術，不易在高溫形成初期有效檢出。此外，鑑於目前國內外相關文獻中，多數皆針對特定高溫形成樣態和模式進行檢測，較缺乏多元類型的有效監測及正確預警的方法，因此本研究發展一整合式防火監診系統，透過整合光纖偵溫及軸承測溫等感測技術與方式，並發展機器學習演算法及建立模型，以有效地降低誤警報，並於輸送帶異況高溫發展的初期，有效地提供預警，以提供使用者迅速掌握資訊，第一時間落實相應措施，避免損害擴大。

關鍵字：輸送機防災、輸送機監診、光纖偵溫、軸承測溫、火災預警。

ABSTRACT

For conveyors, the most concerned and serious disaster is a fire event. Therefore, how to develop a set of methods to effectively achieve the function of early detection and warning has always been the demand of users. However, the causes of fire on conveyor are quite diverse, and it is difficult to detect effectively at the early stage within a single sensing technology. This study develops an integrated fire prevention and monitoring system. Through the proposed methods, such as fiber optic temperature detection and bearing temperature measurement, and the development of machine learning algorithms that can effectively reduce false alarms and provide warnings in the early stage of abnormal high temperature development of the conveyor belt .It can provide user the correct and real time information to avoid damage expansion.

KEYWORDS:Conveyor Monitoring; Distributed Temperature Sensor; Bearing Temperature Detection; Fire Prevention.