應用於銑削監測之嵌入無線溫度感測標籤刀具

Abstract

在追求高精度與永續製造並重的加工產業中，乾銑削因無需冷卻液而具備節能優勢，但也伴隨高溫問題，進而影響刀具壽命與加工品質。傳統接觸式溫度感測技術在高速旋轉與高干擾環境中布設困難，成為刀尖點溫度監測的瓶頸。本研究因此提出一種基於被動式RFID之嵌入無線溫度感測刀具，透過反向散射通訊技術實現非接觸式，且低功耗的被動無線溫度感測系統。 本研究設計並嵌入無源RFID感測標籤於刀具表面，結合熱區隔離策略提升訊號穩定性與感測準確性。透過實驗探討切削深度與進給速率對溫度與訊號穩定性的影響，結果顯示切削深度為主要溫升與通訊干擾來源。為將感測點溫度轉換為刀尖溫度，進一步導入多種機器學習模型進行預測建構，其中時序模型（LSTM、GRU）在處理切削熱的時間相關性上表現尤佳。此外，本研究以數位雙生概念建立設計模擬模型，分別模擬切削力與熱傳行為，並驗證其與實測數據的一致性，有助於探討各種邊界條件下的熱效應與刀具反應，並供未來設計使用。 綜合結果顯示，本系統可於乾銑削加工中穩定量測刀具溫度，並預測刀尖點實際溫度，具備高整合性與擴充性。未來可朝向通訊抗干擾優化、數位模型多物理場耦合發展，進一步應用於永續智慧製造中製程參數的即時量測與調控。

Dry milling has gained attention in modern manufacturing due to its coolant-free, energy-saving nature. However, the resulting thermal accumulation significantly affects tool life and machining quality. Traditional contact-based temperature sensing methods face limitations in high-speed, interference-prone environments, hindering effective thermal monitoring. This study proposes a passive wireless temperature sensing system based on RFID and backscatter communication, embedded into the cutting tool to enable non-contact, low-power temperature measurement. A passive RFID tag was installed on the tool surface, and a thermal isolation strategy was implemented to improve signal stability. Experimental analysis revealed that cutting depth is the dominant factor influencing both temperature rise and signal quality. To estimate tool tip temperature from sensor readings, several machine learning models were applied, with time-series models such as LR and LSTM showing strong performance in capturing thermal dynamics. A digital twin model was also developed using FEM to simulate cutting forces and transient heat transfer, with results aligning closely with experimental data. The proposed system demonstrates stable temperature monitoring during dry milling and accurate tool tip temperature prediction. It offers high integration and scalability, with potential applications in real-time process monitoring and control within sustainable smart manufacturing environments.