Adaptive Wing Morphing Strategy and Flight Control Method of a Morphing Aircraft Based on Reinforcement Learning

基于增强学习的变体飞行器自适应变体策略与飞行控制方法研究

¹ School of Astronautics, Northwestern Polytechnical University, Xi’an 710072, China
² Shanghai Electro-Mechanical Engineering Institute, Shanghai 201109, China

Received: 8 August 2018

Abstract

The morphing aircraft can change different wing shapes or geometries to achieve the optimal flight performance according to various mission scenarios. In this paper, DATCOM is used to calculate aerodynamic parameters based on Firebee UAV morphing aircraft with different wing configurations and analyze aerodynamic characteristics. A novel adaptive wing morphing strategy for morphing aircraft based on reinforcement learning method is proposed. This method can highly meet the demand of keeping optimal performance in multiple flight conditions, and the adaptive wing morphing strategy, three-loop normal load altitude controller and sliding mode velocity controller can together make sure stability of morphing aircraft during morphing process with good tracking performance.

摘要

变体飞行器能根据飞行环境和飞行任务的需要，相应地改变外形，从而始终保持最优的飞行状态，以满足在大飞行包线下执行多种任务的要求。以具有多种翼型的Firebee无人机作为研究对象，利用DATCOM计算气动数据，并展开气动分析。之后，基于增强学习理论，提出一种新型的变体飞行器翼型自适应控制方法。该方法可以很好地满足变体飞行器在多任务状态下保持最优性能的需要，并且设计的高度子系统的三回路法向过载控制器和速度子系统的滑模控制器可以确保飞行器在变体过程中保持稳定，并且跟踪误差较小。