Neural Network Controller Design for a Single Tilt-wing UAV

摘要：

　　This paper presents a design of an adaptive controller for a single tilt-wing UAV using neural networks for full flight stage including hover,transition,and cruise mode.The adaptive controller using the neural network has online adaptation capabilities,which can overcome uncertainties and nonlinearities in system dynamics and the environment.A Dynamic Model Inversion(DMI)is constructed via a six degree of freedom.The DMI controller is used to control inner loop and neural networks compensate for inversion error,which may arise from imperfect modeling,approximate inversion,or sudden changes in the UAV dynamics.The DMI is designed using both non-linear and linear dynamic model.A control performance is verified via numerical simulation.

关键词： Single Tilt-wing Dynamic Modeling Neural-networks Adaptive Controller

作者: Seonghun Yoon Seungkeun Kim Jinyoung Suk

作者单位: Department of Aerospace Engineering,Chungnam National University,Daejeon,Republic of Korea

会议类型: 国际会议

会议名称: 2014 Asia-Pacific International Symposium on Aerospace Technology(2014亚太航空航天技术学术会议)

会议地点: 上海

会议语种:英文

页码: 1-10

在线出版日期: 2014-09-24（万方平台首次上网日期，不代表论文的发表时间）