四旋翼飞行仿真器的PID神经元网络控制器设计  

Design of PID Neural Network Controller for Four Rotor Flight Simulator

作　　者：刘祚时[1] 陈飞 徐燕生 曾锋 LIU Zuo-shi;CHEN Fei;XU Yan-sheng;ZENG Feng(School of Mechanical and Electrical Engineering,Jiangxi University of Science and Technology,Jiangxi Ganzhou 341000,China)

机构地区：[1]江西理工大学机电工程学院,江西赣州341000

出　　处：《机械设计与制造》2023年第1期150-153,158,共5页Machinery Design & Manufacture

基　　金：江西省研究生创新专项资金资助项目(YC2019-S313);江西省重点科技计划项目(20151BBE50038)。

摘　　要：针对四旋翼飞行仿真器系统具有多输入、多输出及多变量等特点,设计了一套基于四旋翼飞行仿真器的PID神经元网络控制器。首先建立四旋翼飞行仿真器系统模型,然后以四旋翼飞行仿真器为控制平台,并以MATLAB/Simulink软件为实验平台,搭建PID控制器和PID神经元网络控制器模型,并分别使用两种控制方法进行实时飞行仿真实验。最后实验结果表明,相比PID控制,PID神经元网络控制方法对四旋翼飞行仿真器的调节时间更短、超调量更小,具有更优的控制性能。APID neural network controller based on the quadrotor flight simulator was designed for the multi-input and multivariable characteristics of the quadrotor flight simulator system.Firstly,the system model of quadrotor flight simulator was established,then,PID controller and PID neural network controller model were built on the control platform of quadrotor simulator and the experiment platform MATLAB/Simulink software and two control methods were used for real-time flight simulation experiments.The final experimental results indicate that,compared with PID control,the PID neural network control method has shorter regulation time and smaller overshoot for the quadrotor flight simulator,and has better control performance.

关 键 词：四旋翼飞行仿真器 控制器 PID神经元网络 实时控制 

分 类 号：TH16[机械工程—机械制造及自动化] TP2[自动化与计算机技术—检测技术与自动化装置]