基于模糊小波神经网络的四旋翼飞行器鲁棒自适应控制

Robust Adaptive Control of Quad- rotor Aircraft Based on Fuzzy Wavelet Neural Network

针对四旋翼飞行器的飞行控制问题,提出了一种基于模糊小波神经网络的鲁棒自适应滑模控制方法。首先利用牛顿-欧拉方程得到四旋翼飞行器的动力学模型,然后用模糊小波神经网络在线逼近系统中的外部扰动和未建模不确定动力学特性,最后提出鲁棒自适应滑模控制算法,通过Lyapunov稳定性理论证明了整个飞行控制系统的稳定性。仿真和实验结果验证了所提出控制方法对外界干扰和模型的不确定性具有较好的鲁棒性和自适应性。

Aiming at the flight control issue of quad- rotor aircrafts,a robust adaptive control approach is presented based on fuzzy wavelet neural network and sliding mode control theory. First of all,the Newton- Euler equation is used to establish the dynamic model of quad- rotor aircrafts,and then by utilizing the fuzzy wavelet neural network to approximate the external disturbance and unmodeled dynamic characteristic in the system,a robust adaptive sliding mode control algorithm is proposed at last.Based on Lyapunov stability theory,the stability of whole flight control system is strictly guaranteed.Moreover,simulation and experiment results validate the proposed control method has superior robustness and adaptiveness for external disturbance and model uncertainty.