本論文以人工呼吸器自動化為主軸，設計具有氧氣回授控制與智慧型呼吸器警報之系統。以呼吸器設計之原則：設定人工呼吸器方式與安全警報兩方面作研究探討。在設定人工呼吸方式部分，所提出的氧氣回授概念，是人工呼吸器自動化中較容易切入研究的一個架構，當病人氧氣需求增加時，由設計的氧氣回授架構，自動依病人需求，利用模糊控制理論，來達到最適切及即時性的氧氣濃度，將醫生的知識、經驗、方法與策略，交由模糊控制器來實現，使得醫學經驗法則容易被理解與接受，其優點是設計上較為簡單與便於應用。而在安全警報部分，本研究開發創新的智慧型警報系統，將呼吸器的生理參數，經綜合演算判斷來達到一較有效率且準確較高的呼吸器警報系統，以提供醫護人員有效的警示及即時處置管理，如病人呼吸器資料收集管理及抽痰、管路脫離或洩漏等各項條件的警示。 最後，本論文已經藉由實驗證實提出之氧氣回授控制與智慧型呼吸器警報判斷，具有其使用優勢與判斷意義在。 This study propose a novel method for mechanical ventilation automation. Based on design principles of ventilation: Setting breathing and security alarms. In the setting the breathing part of the proposed concept of oxygen feedback control, when the patient oxygen requirements increase, the design Oxygen feedback architecture according to patient needs,using fuzzy control theory to achieve the most appropriate and the oxygen concentration in real-time. The alarm system of mechanical ventilation is important to alert caregiver to perform specific procedure to keep patient away from injury and dangerous condition. However, conventional alarm signals are simply designed and need more sophisticated analysis by skilled caregiver to have proper response in short time period. We try to develop an intelligent alarm system to demonstrate more specific clinical problem and allow caregiver to handle the critical situation handy and timely . We set up the alert system to signify the change of lung condition while Pplateau increases. The result presents the suction alarm occurs when Ppeak increase, Rexp increase and Fexp decrease in the mechanical ventilation model after simulated sputum injection.
