多传感器信息融合在自行火炮发动机故障诊断中的应用

Application of Multi-Sensor Information Fusion in Engine Fault Diagnosis of Self-Propelled Gun

针对单一传感器信息反映故障的模糊性和不确定性,发动机故障诊断精度较低的情况,提出了一种基于模糊逻辑的发动机喷油器故障MSF诊断方法。某12缸发动机喷油器上设置的5种故障分别是喷孔堵塞、针阀卡滞、弹簧断裂、针阀开启压力过低和针阀开启压力过高。将8个特征参数融合到模糊逻辑推理诊断系统中,根据实验统计,分别建立了8个特征的隶属度函数和6种状态的输出隶属度函数。该方法能有效地提高发动机故障诊断精度。

Aimed at fuzzy and uncertainty of fault responded by single sensor information, and low accuracy of engine fault diagnosis, MSF fault diagnosis method of engine fuel injector was proposed based on fuzzy logic. A type of engine fuel injector of 12-cylinder was set up to form five kinds of fault, they are nozzle blockage, needle valve clip, spring break, very low and very high pressure of needle valve open. Eight feature parameters were combined with diagnosis system of fuzzy logic inference, and membership degree functions of eight feature parameters and output membership degree functions of six states were set up respectively. This new method can improve fault diagnosis accuracy of engine effectively.