摘要
针对动力翼伞精确数学模型难以获得,系统输入输出关系耦合复杂等特点,建立动力翼伞8自由度动力学模型,设计由静态神经网络和积分器组成的动态神经网络,利用神经网络的逼近能力和动态逆控制方法相结合,提出了基于神经网络动态逆方法的动力翼伞控制方案,并进行了飞行仿真验证,结果表明完全满足控制要求,具有较好的抗干扰能力和鲁棒性能,对于实现动力翼伞的自主飞行控制具有很好的应用价值。
To the problem that the accurate mathematic model of powered paraglider （PPG） can hardly be got, and the flight control system of PPG is a complicated nonlinear system, adynamic model of PPG with 8-DOF is built. A dynamic neural network consisting of the static neural network and the integrators is designed. And the control method of dynamic inversion is used for PPG, based on the ap- pproximation property of neural network and dynamic inversion. Simulation resuhs show that the proposed method strong ability of con- trol and robustness, possessing better application value in autonomous flight control of PPG.
出处
《控制工程》
CSCD
北大核心
2011年第2期178-180,201,共4页
Control Engineering of China
基金
十一五装备预研基金(9140A25050106JB3412)
关键词
动力翼伞
飞行控制
神经网络
动态逆
powerded paraglider
flight control
neural network
dynamic inversion
