摘要
考虑甲板运动和舰尾流等扰动影响下的舰载机着舰轨迹跟踪问题,提出一种基于预定义时间的自适应抗干扰控制策略.建立着舰轨迹生成、引导、控制和进近动力补偿等子系统,将轨迹跟踪问题转换为子系统的稳定问题.针对舰尾流引起的时变扰动和甲板运动对理想着舰点的变化影响,采用长短期记忆神经网络进行甲板运动预估并在引导指令中予以修正,借助非线性扰动观测器估计未知干扰对模型集总外界扰动进行前馈补偿.为提升着舰轨迹跟踪与姿态控制的精确性和快速性,设计基于反步架构的预定义时间控制策略,通过李雅普诺夫稳定性分析证明系统能够在设定的时间内收敛.数字和半实物仿真结果表明,在甲板运动和舰尾流扰动影响下所设计的控制策略能够满足着舰轨迹的快速准确跟踪,实现预定义时间稳定.
Considering the landing trajectory tracking of carrier-based aircraft under the influence of deck motion and ship wake,an adaptive robust control strategy based on predefined-time is proposed.The subsystems for ship trajectory generation,guidance,control,and approach power compensation are established,transforming the trajectory tracking problem into a stability problem for these subsystems.To address the influence of time-varying disturbances caused by ship wake and deck motion on the ideal landing point,a long short-term memory neural network is utilized for deck motion estimation and correct it in the guiding commands,and a nonlinear disturbance observer is designed to estimate unknown disturbances so that the feedforward compensation is performed on the aggregate external disturbance of the model set.To enhance the accuracy and speed of landing trajectory tracking and attitude control,a predefined-time control strategy based on backstepping architecture has been developed.Lyapunov stability analysis demonstrates that the system can converge within the set time.The numerical and hardware-inloop simulation results indicate that the control strategy designed to account for deck motion and ship wake disturbances can meet the requirements for fast and accurate tracking of landing trajectories while achieving predefined-time stability.
作者
李钊星
蔡云鹏
刘茂汉
王霞
许斌
LI Zhao-Xing;CAI Yun-Peng;LIU Mao-Han;WANG Xia;XU Bin(University,Xi'an 710072;Shenyang Aircraft Design&Research Institute,Shenyang 110035;School of Control Science and Engineering,Shandong University,Jinan 250061;Shenzhen Research Institute of Northwestern Polytechnical University,Shenzhen 518057)
出处
《自动化学报》
北大核心
2025年第6期1233-1247,共15页
Acta Automatica Sinica
基金
陕西省自然科学基础研究计划(2023JC-XJ-08)
深圳市科技计划(JCYJ20230807145500002)
西北工业大学博士论文创新基金(CX2024071)资助。
关键词
着舰控制
舰尾流
甲板运动
长短期记忆神经网络
预定义时间控制
Landing control
ship wake
deck motion
long short-term memory neural network
predefined-time control
