基于Adaline网的航空发动机自适应控制

Aeroengine adaptive control based on adaline network

关键词:  航空发动机  人工神经元网络  神经网络辨识器  神经网络控制器  自适应控制

Key words:  Aircraft engine  Artificial neural network  Neural estimator  Neural controller  Adaptive control

摘要:

根据发动机模型非线性、关系式呈隐性和具有不确定性的工作特点 ,提出了一种基于Adaline网的发动机自适应控制方案 ,该方案中的神经网络辨识器和神经网络控制器均采用两层线性Adaline网 ,权值采用Widrow Hoffδ学习规则更新。仿真结果表明这种方法具有方案简单、稳定性好、实时性好等优点 ,可在全飞行包线内对发动机进行在线控制

Abstract:

Associated with the characteristic of nonlinear,recessive and uncertain of model,a kind of aeroengine adaptive control scheme based on adaline network was presented.Two layered linear adaline network were used both in the neural estimator and nueral controller.The weight was updated using Widrow Hoff delta rule. Simulation results show that the scheme has the advantage of briefness,stability and better real time ability.It can be used on the on line control of aeroengine throughout the full flight envelope.