基于神经网络的飞艇非线性鲁棒滑模飞行控制

Nonlinear robust sliding-mode flight control for the airship based on neural networks

摘要:

为了提高飞艇飞行的鲁棒性能，针对飞艇水平面受扰运动提出了一种基于非线性滑模的鲁棒神经网络飞行控制方案。对于飞艇模型中的不确定性，采用神经网络进行逼近，同时对神经网络权值进行在线调整以达到在线逼近的目的。为了有效地处理外部时变干扰对飞艇飞行控制的影响，将参数自适应、神经网络输出与滑模控制相结合设计了飞艇的鲁棒飞行控制。最后，通过飞艇水平面运动控制验证了本文所提出的飞艇鲁棒飞行控制方法的有效性。

Abstract:

The robust neural network flight control scheme is proposed for the plane movement of the airship with disturbance based on the nonlinear sliding mode technology to improve the flight robust performance. The neural networks are used to approximate the uncertainty of the airship model and the weight values are adjusted on line to achieve the approximation. To effectively handle the external disturbance of the airship，the robust flight control is designed combining the parameter adaptation with neural networks and sliding mode control for the airship. Finally，the plane movement control of the airship is executed to demonstrate the feasibility of the proposed robust airship flight control method.