基于新型神经网络的航空发动机多变量控制

根据PID控制结构提出了一种新型神经网络控制器,对其基本结构和学习算法等进行了分析。结合某型航空涡喷发动机双变量控制需求,利用2个结构相同、相互联系的神经网络,实现了发动机双变量控制和接耦。在不同飞行条件、不同发动机工作状态下的仿真表明,控制器具有良好的控制性能

Aircraft Engine Double- Variable Controller Based on a New Neural Network

A new neural network controller is proposed based on the PID controller structure.Its basic structures and learning algorithm are analysed.For an aircraft turbojet engine,the double-variable control and decoupling are accomplished through two interconnected neural networks with the same structure.The simulation results obtained at different flight and operating conditions have shown the perfect performance of the controller.