论文部分内容阅读
针对航空发动机多变量控制系统中各回路之间存在的耦合现象,提出了一种基于RBF网络辨识的航空发动机多变量单神经元网络解耦控制方法。对发动机的多个控制回路,采用多个RBF网络实时辨识各个回路发动机的数学模型,并将系统的灵敏度信息实时反馈给各回路的控制器,保证了单神经元网络控制器对各回路的准确控制,最终实现对发动机多回路的解耦控制。通过在飞行包线内的仿真,结果表明,该方法不依赖被控对象的精确模型,有效地实现了对发动机的解耦控制,而且具有良好的动静态性能,将其应用于航空发动机多变量解耦控制是行之有效的。
Aiming at the coupling phenomenon existing in each loop of aero-engine multivariable control system, a multivariable single-neuron network decoupling control method based on RBF network identification is proposed. For multiple control loops of the engine, multiple RBF networks are used to identify the mathematical models of each loop engine in real time, and the sensitivity information of the system is fed back to the controllers of the loops in real time to ensure that the single neural network controller is accurate for each loop Control, the final realization of the engine multi-loop decoupling control. The simulation results show that this method does not depend on the precise model of the controlled object and effectively realizes the decoupling control of the engine, and has good dynamic and static performance, which is applied to aero-engine multivariable Decoupling control works well.