基于免疫理论的航空发动机磨损故障智能诊断

针对航空发动机磨损故障诊断技术智能化、精确化的发展要求,以传统油液监测技术为基础,结合人工免疫系统(AIS)具有的自适应特性、学习记忆特性及识别特性等优点,提出了一种航空发动机磨损故障的智能诊断方法。该方法首先利用人工免疫理论的反面选择原理生成检测器,优化后的检测器生成算法提高了初始检测器的代表性及覆盖性;然后利用故障样本训练出成熟的检测器,使航空发动机典型的磨损状态信息存储在检测器中,实现对故障模式的有效学习和记忆;最后通过检测器的激活发现航空发动机的磨损故障。对油样数据的实例分析结果表明,该方法对航空发动机磨损故障具有较强的识别能力,对磨损状态有很好的监测效果。

Intelligent diagnosis for aircraft engine wear fault based on immune theory

To meet the development requirements of intelligentization and accuracy for the aircraft engine wear fault diagnosis technology, an intelligent diagnosis method for aircraft engine wear fault is proposed based on the traditional oil monitoring technology, and this method is combined with the artificial immune system's (AIS) advantages, such as adaptive characteristic, learning and memory characteristic and recognition characteristics. Firstly, the method uses negative selection principle of artificial immune theory to build detectors. The optimized immune algorithm improves the initial detectors' representativeness and coverage. And then fault samples are used to train and generate mature detectors. So the typical information of aircraft engine wear fault is stored in the mature detectors, implementing effective learning and memory of the failure modes. Finally, wear fault of the aircraft engine can be found through the activated detectors. The case analytical results of the sample data demonstrate that the method has strong ability to recognize aircraft engine wear faults and it has very good monitoring effect to wear condition.