飞机压力加油系统实时传感器的信息有效性分析

Real Time Sensor Validation for Airplane Pressure Fueling System

针对飞机地面压力加油系统传感器的实测信息 ,提出了实际环境下的传感器信息确认算法 .利用贝叶斯网络 (BN)构造传感器的相互关系 ,描述了各传感器之间相互依赖与独立的程度 ,并且对传感器参数进行动态评估 ,得出了对系统状态的准确判断 ,用于指导对故障的有效确认 .在地面试验研究过程中 ,对一起管路严重超压现象进行了成功的判断 ,证明了该方法的可行性和有效性 .

Aimed at the on-line data from the 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 judgments to the system state can be deduced to guide the fault isolation. During the experiments, an incident of pressure overshoot comes forth in the entrance pipeline connected with the reservoir No.5, and the analytical result shows something wrong with the corresponding oil height sensor. According to the phenomenon that the inner resistance mismatch really occurs in the same sensor, the experiment strongly testifies that the method described above is feasible and efficient.