基于模糊神经网络的涡喷发动机控制系统设计  

Design of Turbojet Engine Control System Based on Fuzzy Neural Network

作　　者：李慧琳 封锋[1] Li Huilin;Feng Feng(School of Mechanical Engineering,NUST,Nanjing 210094,China)

机构地区：[1]南京理工大学机械工程学院,南京210094

出　　处：《计算机测量与控制》2021年第2期53-57,共5页Computer Measurement &Control

摘　　要：针对微型涡喷发动机ECU控制系统具有时变性和非线性的特点,为改善微型涡喷发动机控制系统的控制性能,将模糊神经网络PID控制方法应用于ECU的转速与推力控制系统中;首先,利用某微型涡喷发动机的试车数据通过系统辨识方法得到其数学模型,其次针对模糊PID无法在线调参的弊端,引入模糊神经网络控制方法对微型涡喷发动机ECU系统进行控制;为模拟发动机在工作过程中遇到的干扰问题,在仿真过程中加入了干扰信号,通过与传统PID、模糊PID的仿真结果对比验证得出,模糊神经网络PID在涡喷发动机转速控制系统中响应速度更快约为1 s,超调量更小约为0,在有干扰的情况下恢复稳定状态的时间更短,约为0.5 s。In view of the characteristics of time-varying and non-linear characteristics of the ECU control system of the micro-turbojet engine,in order to improve the control performance of the micro-turbojet engine control system,the fuzzy neural network PID control method is applied to the speed and thrust control system of the ECU.First,using the test data of a micro-turbojet engine to obtain its mathematical model through the system identification method,and secondly,the fuzzy neural network control method is introduced to control the micro-turbojet engine ECU system for the drawback of fuzzy PID online parameter adjustment.In order to simulate the interference problems encountered by the engine in the process of operation,interference signals were added to the simulation process.Through comparison and verification with the simulation results of traditional PID and fuzzy PID,it is found that the fuzzy neural network PID control system is better.In the turbojet engine speed control system,the fuzzy neural network PID has a faster response speed of about 1 s,a smaller overshoot of about 0,and a shorter time to return to a stable state in the presence of interference,about 0.5 s.

关 键 词：微型涡喷发动机 模糊神经网络控制 ECU控制系统 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]