论文部分内容阅读
首先提取航空发动机排气温度数据序列的边界值,并且证明了边界值序列具有混沌特征.其次提出了一种基于多元相空间重构的发动机状态混沌预测算法,实现对排气温度的预测.通过检验排气温度预测值是否超过所规定的红线,从而进行发动机的健康状态排查.作为验证实例,使用一组某机型发动机实际飞行数据对该预测算法进行了验证,结果表明:该组合算法降低了预测模型的时间复杂度,并大大提高了疑似异常点的预测精度.该方法可以为这种机型发动机故障预测提供决策依据.
Firstly, the boundary value of the aero-engine exhaust temperature data sequence is extracted, and the chaotic characteristics of the boundary value sequence are proved.Secondly, a state chaos prediction algorithm based on multiple phase space reconstruction is proposed to predict the exhaust gas temperature. Test the exhaust temperature predicted value exceeds the specified red line, so as to carry out engine health check.As a verification example, the actual flight data of a certain type of engine was used to verify the prediction algorithm, the results show that the combination algorithm reduces The time complexity of the prediction model is raised and the prediction accuracy of the suspected anomalous point is greatly increased.This method can provide decision basis for the engine fault prediction of this model.