由於直升機充分仰賴供其完成各種飛行動作之機械飛操控制以及高科技裝置，使得故障診斷與排除工作亦增加其困難與複雜性。誤判故障極難避免，也因此相對地增加維修及維護成本。雖然資深的修護人員能克服此問題，然而仍需借助其豐富的修護經驗，且往往費力耗時。隨著資訊科技的日益成熟，藉由萃取知識的過程儼然成為軍隊提昇戰力的重要關建因素，為克服直升機故障診斷瓶頸，本研究結合資訊科技，應用案例式推論與規則式推論方法，發展直升機故障診斷與排除專家系統供決策輔助，並強化系統學習的功能將維修知識回饋至案例資料庫，透過此資料庫將資訊分享給所有修護人員，以避免在直升機修護運作過程發生故障誤判情事。實驗結果，應用本研究推論引擎所建議之方法分析故障，能維持更佳之直升機修護品質與效能。除此之外，也可藉由此系統做為人員教育訓練教材，增加經驗累積，達到知識管理的目的。 The complicated flight control mechanisms and high-tech devices composed in a helicopter are often the reasons for perplexed fault diagnosis and isolation for a helicopter. An erroneous judgment may happen thus increasing the maintenance costs. Although a skilled technician can uncover the possible defect, this time consuming approach requires strong engineering experience. In order to solve this problem, this research applies both case-based reasoning (CBR) and rule-based reasoning (RBR) methodology to develop a fault diagnosis and isolation system (ESHFDI) to serve as a decision aid to avoid misjudgments in the helicopter maintenance operation. According to our experimental results, applying our proposed method in the inference engine to analyze failure improves the helicopter maintenance quality.
