通过对航空发动机的典型故障分析,应用RBF神经网络,构建不同油门开度下的发动机性能参数,进行故障特性学习作为训练样本,将测试样本与训练样本的期望值进行比较,来确定发动机性能的衰变程度,提出了航空发动机性能评价模型.通过多机种多架次的大量飞行记录数据的实验,验证了该方法是能够及时检测到故障的发生和识别的一种有效方法.通过此方法能够成功地对发动机的健康状态进行早期诊断与预报.%Through the analysis of typical aeroengine fault, using RBF neural network, this page constructs performance parameters under the throttle angle of engine and studies the fault characteristics as the training sample. Comparing with an expected value between test samples and training samples, we can get the decay degree of engine performance and the aeroengine performance evaluating model. By large amounts of aeroengine recorded flight data tested, plenty of other vehicles validated that this method can diagnose engine fault accurately and timely. It can diagnose and predict aeroengine health status successfully.
展开▼