提出一种基于邻域粗糙集和支持向量机相结合的航空发电机智能健康诊断方法。采用专业健康试验平台对某型战斗机的真实航空发电机进行试验,得到转速、负载、油压等大量表征发电机健康状态的监测数据。引入数据挖掘思想,采用邻域粗糙集理论对监测数据进行属性约简,将约简后的属性集输入给所设计的支持向量机健康诊断器,对航空发电机的健康状态进行了诊断研究。研究表明,该方法能够很好实现对某真实航空发电机的健康诊断,具有较高的推广应用价值。 An intelligent health diagnosis method based on neighborhood rough set and support vector machine is proposed. A professional health test platform was used to test a real aircraft generator of a fighter plane. The monitoring data of the generator health status, such as speed, load and hydraulic pressure, were obtained. The idea of ​​data mining was introduced. The neighborhood rough set theory was used to reduce the attribute of the monitoring data. The reduced attribute set was input to the designed SVM health diagnosis device, and the health status of the aircraft generator was diagnosed. The research shows that this method can well realize the health diagnosis of a real aero-generator, and has a high popularization and application value.