基于遗传算法和BP网络的航空发动机拆换期望值预测

通过使用真实样本实验的方法,在BP神经网络、GA遗传算法与改进的GA-BP复合算法中,找出能迅速精确地预测航空发动机拆换期望值的最佳方法。试验结果证明,GA-BP复合算法在用遗传算法对神经网络的权值进行大致搜索以后,再用神经网络方法进行训练,能很好地模拟发动机拆换期望值,并用实例证明该算法是有效的。 

The Estimation of Aeroengine Expectant Cycle Based on Genetic Algorithm and BP Neural Network

The algorithm described in the paper is one of the important instruments for the establishment of aero engine expectant cycle to supervise aero engine scientifically and effectively.By means of experimental method of using real specimens,the optinal algorithm which can estimate aero engine expectant cycle quickly and precisely among BP neural network,GA (genetic algorithm) and compound GA BP can be found.It shows that GA BP algorithm can simulate expectant cycle excellently.First it searches weights of neural network grossly using genetic algorithm,then is trained by using BP network.It has been proved to be effective by real instances.Two softwares have been made.