基于QDRNN网络的航空发动机多变量解耦控制
Multivariable decoupling control of aeroengine based on QDRNN
-
摘要: 针对航空发动机非线性模型的复杂性,通过对准对角递归神经网络(QDRNN)及梯度下降法(GMD)的分析,研究了基于QDRNN网络的航空发动机多变量解耦PID控制系统.阐明了该算法的结构和原理,通过对设计点及非设计点的仿真.研究表明,QDRNN网络结构相对简单,易于构造训练算法,较好地解决了双变量控制系统中变量之间的耦合作用.
-
关键词:
- 航空、航天推进系统 /
- 航空发动机 /
- 准对角递归神经网络(QDRNN) /
- 多变量控制 /
- 解耦
Abstract: Through analysis of QDRNN(quasi-diagonal recurrent neural network) and GMD(gradient decent method),a multivariable decoupling PID(proportional integral,differential) control system based on QDRNN was presented to resolve the complex nonlinear model of aeroengine.The structure and principle of the algorithm were also illustrated through simulation of design points and non-design points.The research results show that,a multivariable decoupling PIDcontrol system based on QDRNN is well-suited for decoupling of variables in dual-variable control system. -
[1] Kumpati S N,Ukhopadhyay S M.Intelligent control using neural networks[J].IEEE Control,April,1992:11-18. [2] Narendra K S,Parthasarathy K.Identification and control for dynamic systems using neural networks[J].IEEE Trans.On Neural Networks,1990,1 (1):4-27. [3] Chen T C,Sheu T T.Model reference neural network controller for induction motor speed control[J].IEEE Transctionson Energy Conversion,2002,17 (2):157-163. [4] Chow W S,Fang Y.A recurrent neural-network-based real-time learning control strategy applying to nonlinear systems with unknown dynamics[J].IEEE Trans on Neural Network,1998,45(1):151-161. [5] Dray J S.Dynamic recurrent neural networks; A dynamical analysis[J].IEEE Transaction on Systems,Man,and Cybernetics,Part B:Cybernetics,1996,26(5):692-706. [6] 李鸿儒,王建辉,顾树生.准对角递归神经网络及其算法的研究[J].系统仿真学报,2004,07:1542-1544.LI Hongru,WANG Jianhui,GU Shusheng.Study on a quasi-diagonal recurrent neural network and Its algorithm[J].Acta Simulata Systematica Sinica,2004,07:1542-1544. [7] 刘金琨.先进PID控制及其MATLAB仿真[M].北京:电子工业出版社,2003.
点击查看大图
计量
- 文章访问数: 1317
- HTML浏览量: 2
- PDF量: 389
- 被引次数: 0