針對飛機地面壓力加油系統傳感器的實測信息,提出了實際環境下的傳感器信息確認算法.
利用貝葉斯網絡(BN) 構造傳感器的相互關系,描述了各傳感器之間相互依賴與獨立的程度,并且對傳感器參數進行動態評估,得出了對系統狀態的準確判斷,用于指導對故障的有效確認. 在地面試驗研究過程中,對一起管路嚴重超壓現象進行了成功的判斷,證明了該方法的可行性和有效性.
關鍵詞: 壓力加油系統;傳感器;貝葉斯網絡;有效性
Abstract : Aimed at the on - line data f rom sensors of airplane pressure fueling system , an algorithm for sensor validation is presented with the help of Bayesian network (BN) . The degrees of dependence and independence among all the sensors are determined. Then the accurate judgment s to the system state can be deduced to guide the fault isolation. During the experiment s , an incident of pressure overshoot comes forth in the ent rance pipeline connected with the reservoir No. 5 , and the analytical result shows something wrong with the corresponding oil height sensor. Through the succedent detection , the phenomenone that the inner resistance mismatch really occurs in the same sensor , the experiment testifies st rongly that the method described above is feasible and efficient .
Keywords : pressure f ueling system ; sensors ; Bayesian network ; validation