为进一步减小排气温度裕度计算误差,对发动机起飞排气温度裕度基线观察值和雷诺数影响系数进行了多元非线性拟合,提出了利用雷诺数影响系数修正排气温度(Exhaust Gas Temperature,EGT)基线观察值的方法,将雷诺数影响系数加入神经网络的输入层,利用遗传算法(Genetic Algorithm,GA)优化Elman网络模型,建立排气温度裕度(Exhaust Gas Temperature Margin,EGTM)的预测模型.通过结合飞行数据计算,对比多元非线性拟合以及Elman网络模型和基于Elman网络优化的GA-Elman模型的计算误差效果,得出实验结果:GA-Elman对EGTM计算精度更高,鲁棒性更强.
Research of the Corrected GA-Elman Algorithm with the Reynolds Number on EGTM
In order to further reduce the calculation error of exhaust gas temperature margin(EGTM),the baseline observation of the take-off EGTM and the influence coefficient of the Reynolds number of the engine are carried out multivariate nonlinear fitting,and a method of correcting the baseline observation of exhaust gas temperature(EGT)by using the influence coefficient of the Reynolds number is proposed.The influence coefficient of the Reynolds number is added to the input layer of the neural network,the Elman network model is optimized by the genetic algorithm(GA),and the prediction model of EGTM is established.By combing with flight data calculation,the calculation error effects of multivariate nonlinear fitting and the Elman network model and the GA-Elman model based on Elman network optimization are compared,and the experimental results show that GA-Elman has higher calculation accuracy and stronger robustness to EGTM.
AeroengineExhaust gas temperature marginReynolds numberBaseline observationGenetic algorithm
NETL
NSTL
万方数据
维普
