论文部分内容阅读
提出了一种发动机多变量神经网络控制方法。采用三层前向神经网络对非线性项进行补偿，神经网络权重自适应律采用在线学习算法。控制系统的设计不需要知道精确模型，适应于全飞行包线。仿真研究表明发动机多变量神经网络控制具有良好的动、静态性能。
A multivariable neural network control method is proposed. The three-layer feedforward neural network is used to compensate the nonlinear term, and the adaptive algorithm of neural network weight is based on online learning algorithm. Control system design does not need to know the exact model, adapted to the full flight envelope. Simulation studies show that the multivariable neural network control of the engine has good dynamic and static performance.