Predictive maintenance (PM) strategies are based on real-time data for diagnosis of impending failure and prognosis of machine health. It is a proactive process, which needs predictive modeling to trigger an alarm for maintenance activities and anticipate a failure before it occurs. Various industries have adopted PM techniques because of its advantage in increasing reliability and safety. But in the aviation industry, expectations for safety are increased due to its high cost and danger to human life when an aircraft fails or becomes out of service. Flight data monitoring systems are regularly implemented in commercial operations using artificial intelligence (AI) algorithms, but there is limited work specific to safety critical systems such as engine and hydraulic system. This paper provides a survey of recent work on PM of aircraft's’ hydraulic system and engine, identifying new trends and challenges. This work also highlights the importance of PM and state-of-the-art data pre-processing techniques for large datasets.

预测性维护 (PM) 策略基于实时数据，用于诊断即将发生的故障和预测机器健康状况。这是一个主动过程，需要预测建模来触发维护活动警报并在故障发生之前预测故障。由于 PM 技术在提高可靠性和安全性方面的优势，各个行业都采用了 PM 技术。但在航空业，由于成本高昂以及当飞机出现故障或停止服务时对人的生命构成危险，因此对安全性的期望越来越高。使用人工智能 (AI) 算法在商业运营中定期实施飞行数据监控系统，但针对发动机和液压系统等安全关键系统的工作有限。本文对近期飞机 PM 相关工作进行了综述。s 的液压系统和发动机，识别新的趋势和挑战。这项工作还强调了 PM 和最先进的数据预处理技术对大型数据集的重要性。