基于滑模神经网络的自主飞艇姿态控制  

Control System of Autonomous Airship Based on Siding Mode Control and Neural Network

作　　者：郭建国[1] 周军[1] 

机构地区：[1]西北工业大学精确制导与控制研究所,陕西西安710072

出　　处：《飞行力学》2009年第1期40-42,共3页Flight Dynamics

基　　金：国家高技术发展计划资助项目(2007HG20016)

摘　　要：针对自主飞艇飞行环境的不确定性,提出了一种基于自适应滑模神经网络的姿态控制系统。平流层高空飞行环境对飞艇控制产生了许多不确定性因素,利用自适应变结构控制和神经网络方法设计了飞艇的俯仰通道控制器。非线性仿真结果表明:控制器能够适应对象结构参数及外部扰动的大范围变化,满足姿态控制稳定性要求,同时也消除了变结构控制系统的抖振,具有良好的鲁棒性和动态性能。An adaptive controller combining the siding mode control with neural network is derived for autonomous airship according to flight environment. Due to uncertainties of stratosphere, the control system of pitch channel was designed based on adaptive variable structure control and neural network method. By the nonlinear simulation, it was showed that the controller could adapt well to large-scale variation of model parameter perturbation and wind turbulence which could meet requirement of control system stability, and the chattering of VSC was eliminated. The method made the control system good robustness and dynamic performance.

关 键 词：自主飞艇 变结构控制 神经网络 飞艇飞行动力学 

分 类 号：V249.1[航空宇航科学与技术—飞行器设计]