基于边界值的多元混沌发动机性能预测算法

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. 

Multivariable chaotic arithmetic for aero-engine performance forecasting based on boundary values

Firstly,the boundary values of exhaust gas temperature(t_eg) was extracted from the original exhaust gas temperature data series.Further analysis on the boundary values data series indicated that there existed a chaos feature in it.Then,chaotic forecasting arithmetic based on multivariable phase space reconstruction was established to forecast the data series.Finally,by comparing the series with the red line,the condition of aero-engine was defined.The proposed arithmetic was verified through some type of aircraft aero-engine t_eg data series obtained from actual flight.The result shows that the proposed arithmetic can reduce the time complexity and has a better forecast accuracy in the prediction of outliers.It can be used as a supporting model in the decision making of this aero-engine fault forecast.