基于模糊神经网络PID控制器的EEC数字化设计

Digitization Design of EEC Based on Fuzzy Neural Network PID Controller

为实现航空发动机模拟式电子控制器(EEC)的数字化设计,以其低压压气机导流叶片调节通道为主要研究对象,提出一种模糊神经网络PID控制器,将模糊控制、神经网络、PID控制相结合,利用模糊控制专家经验优势和神经网络的自学习、自适应能力,优化PID控制参数,实现控制性能提升。仿真结果显示,基于模糊神经网络的PID控制器控制性能有较大提高,具有比常规神经网络PID控制器更小的超调量和更好的抗干扰性;适用于定常系统和非定常系统,具有更好的自适应性与鲁棒性;可应用于航空发动机模拟式电子控制器(EEC)的数字化设计。

In order to realize the digitization design of the analog electronic engine controller （ EEC）, the inlet guide vane angle adjusted by the low pressure compressor is used as the main study object. A fuzzy neural network PID controller combining fuzzy control, neural network and PID control is proposed, the fuzzy control expert experience and the self-learning and adaptive ability of neural network are also utilized for optimization of PID control parameters to improve the control pefformance. The simulation results show that the PID controller based on fuzzy neural network has a better control pefformance, and it has a smaller overshoot than the conventional neural network PID controller, and better anti-interference ability. It is suitable for both steady and unsteady systems with better adaptability and robustness. The controller can be applied to the digital design of the analog EEC.