神经网络在发动机状态监控与故障诊断中的应用综述

神经网络由于具有强大的非线性映射能力和并行处理能力,近年来,在航空航天领域中,被广泛地应用于发动机的状态监控和故障诊断.全面总结了神经网络在发动机状态监视与控制、故障诊断以及可靠性分析与设计上的应用情况,系统地分析了所采用的反向传播神经网络、径向基函数神经网络、自组织神经网络、概率神经网络以及模糊神经网络等网络模型及其最新的进展情况,最后指出了未来的重点研究方向.

A Review of Neural Network Applied to Engine Status Monitoring and Fault Diagnosis

The neural network has the powerful ability of non-linear map ping and parallel processing; therefore, during the recent years , it is widely applied to status monitoring and fault diagnosis of engine in the aeronautical a nd astronomical fields. In the paper the applications of networks to the status monitoring, the fault diagnosis, and reliability analysis and design are all con cluded. Then the models of back propagation, radical basis function, self-organ izing, probabilistic, and fuzzy neural networks and their up-to-date developme nts are studied and analyzed respectively. Lastly, some important fields for fut ure research are presented.