航空电子设备故障预测框架与方法

针对环境应力对电子产品造成的累积损伤,提出了覆盖电子设备全寿命周期的故障预测框架与方法。首先提出了可预测性的概念,在设备设计阶段采用辅助软件eXpress进行可预测性设计,跟踪实际环境中设备的工作数据,采用3种改进神经网络进行特征参数的趋势预测。基于信息熵理论,采用多指标评价准则融合单项预测方法,实现了对某型电子设备的故障组合预测。预测框架在某型电子设备上的实践验证了框架的可行性,预测结果表明了组合预测方法较单项预测方法具有更强的鲁棒性和稳定性。

Prognostic Framework and Approach for Electronic Equipments

To treat with accumulative damage brought by environmental stress, a prognostic framework and a prediction method for electronic equipments are proposed, which cover the whole life cycle of the electronic equipments. The conception of prognostic capability is put forward for the first time. The design for prognostic ability is carried out by adopting auxiliary software eXpress at the design phrase of equipment. With the working data of equipment in practical environment, three improved neural networks are studied to predict the developing trend of characteristic parameters. By fusing single prediction methods with multi-rules, one combined fault prediction method based on information entropy theory is proposed to enhance the reliability of prediction results. The application of prognostic framework to one type of electronic equipment has verified the feasibility of the framework. And the prediction results show that the combined prediction method is stronger than single prediction methods in robust ability and stability.