摘要
频域内的神经控制方法在振动主动控制中因其允许复杂的控制算法 ,适合于智能旋翼的控制。在建立频域智能旋翼系统模型的基础上 ,提出了频域内的智能旋翼神经控制方法。采用正交设计方法选取训练样本 ,建立频域神经模拟器 ,然后按照提出的算法进行控制。仿真结果表明提出的算法有效 ,并且具有较好的鲁棒性。
The smart rotor, a hot research point in the field of helicopter engineering in recent years, has the characteristic of nonlinearity and periodicity and is suited to be researched in the frequency domain. Because of the permissibility of the complicated control algorithm in active vibration control, the neural control method in the frequency domain is suitable for control of a smart rotor. Based on the smart rotor system modeled, the neural control method in the frequency domain for the smart rotor is put forward. The neural emulator in the frequency domain is advanced with the training cases selected by an orthogonal design method. Calculation results show that the availability and robustness of the neural control method in the frequency domain.
出处
《航空学报》
EI
CAS
CSCD
北大核心
2001年第5期420-423,共4页
Acta Aeronautica et Astronautica Sinica
基金
国家自然科学基金 ( 5 9875 0 35 )
教育部博士点基金( 19990 2 870 6 )资助项目
关键词
振动主动控制
智能旋翼
神经网络
附翼
频域
直升机
Algorithms
Frequency domain analysis
Neural networks
Nonlinear control systems
Robustness (control systems)
Rotors
