大型飞机发动机故障智能诊断方法研究与仿真

摘要

Research failure accurate recognition of large aircraft engine. Large aircraft engine structure is complex and has many noises. This paper put forward a fault diagnosis methods based on particle swarm rough intensive Jane. Through the fault knowledge acquisition and rules optimization, the fault prediction method was used to complete fault identification. The experiment results show that the algorithm improves the accuracy of identification.%研究大型飞机发动机中故障准确诊断的问题.与一般元器件不同,大型飞机发动机结构复杂,工作中出现故障频率较高,故障出现时间较短,并且很难复现,导致在建立故障间的关联时出现信息关联断裂.传统的故障检测方法多是根据故障信息之间的关联性进行判断,一旦故障不能复现,将导致判断失败,降低了检测的准确性.提出了粒子群粗糙集约简的故障诊断方法,并且应用到飞机发动机中,通过对故障知识获取、规则优化,运用故障预测的手段,重建故障信息断裂的关联性.完成故障识别.实验结果证明,这种方法较好地提高了故障诊断的准确性.