基于SSA-LSTM的高高原民航飞机空调系统故障预测

Fault Prediction of High-Altitude Civil Aviation Aircraft Air Conditioning System Based on SSA-LSTM

在民航高高原航线的复杂气候环境下,机载设备整体性能要求更高。在低温、低压、强紫外线等条件下,飞机关键部件会加速老化。国内航空公司通常采用事后维修的方式对飞机进行维护,针对传统事后维修方法响应时间长、成本高以及存在潜在的安全风险,难以人为精准把控维修周期和维修深度的问题,提出一种大数据驱动的飞机空调系统故障预测方法。引入融合麻雀算法的长短期记忆网络(Sparrow Search Algorithm-Long Short-Term Memory,SSA-LSTM),利用无线快速访问记录器(Wireless Quick Access Recorder,WQAR)收集的数据,与其余两种预测方法进行对比研究,结果表明SSA-LSTM具有明显优势。通过预测和早期潜在故障识别,为航空公司从事后维修向预防性维修的转变提供了支持。

In the complex climatic environment of high-altitude airline,the overall performance requirements for onboard equipment are higher.Under conditions of low temperature,low pressure,and strong ultraviolet radiation,the aging of critical aircraft components is accelerated.Domestic airlines usually adopt a post-maintenance approach to aircraft upkeep.In response to the traditional post-maintenance method,which involves long response times,high costs,and potential safety risks,and which makes it difficult to manually control the maintenance cycle and depth accurately,a big data-driven method for predicting failures in aircraft air conditioning systems is proposed.This method introduces the Sparrow Search Algorithm-Long Short-Term Memory(SSA-LSTM)network.Utilizing data collected by the Wireless Quick Access Recorder(WQAR),and comparing it with two other prediction methods,the results demonstrate a clear advantage of the SSA-LSTM.By predicting and identifying potential early faults,this method supports the shift from post-maintenance to preventive maintenance for airlines.