A neural approximation-based novel back-stepping control scheme for air-breathing hypersonic vehicles with uncertain parameters. (March 2016)

• Record Type: Journal Article
• Authors: Bu, Xiangwei; Wu, Xiaoyan; Zhang, Rui; Ma, Zhen; Huang, Jiaqi
• Is Part Of: Proceedings of the Institution of Mechanical Engineers. Volume 230:Number 3(2016)
• Page Start: 231
• Page End: 243
• Publication Date: 2016-03
• Subjects: Air-breathing hypersonic vehicles -- parameter uncertainties -- improved back-stepping -- minimal-learning-parameter -- neural approximation; Mechanical engineering -- Periodicals; Automatic control -- Periodicals; Systems engineering -- Periodicals; 621.3
• DOI: 10.1177/0959651815622329
• Languages: English
• ISSNs: 0959-6518

Abstract

In this article, a novel back-stepping control strategy is proposed for the longitudinal dynamics of air-breathing hypersonic vehicles subject to parameter uncertainties based on neural approximation. To facilitate the control design, the vehicle dynamics is reasonably decomposed into subsystems including the altitude subsystem and the velocity subsystem. Different from the existing studies, the presented improved back-stepping control approach for altitude dynamics only contains one actual controller while all the virtual controllers are artificial intermediate states needed only for stability analysis purpose. Hence, the problem of "explosion of terms" is completely avoided. By utilizing neural networks to approximate the lumped uncertainty of each subsystem, the robustness of the explored controller against uncertain aerodynamic coefficients is guaranteed. Moreover, by the merit of the minimal-learning-parameter method, only one learning parameter is required to be updated in each subsystem. The novelty of this article is that the exploited control scheme is simplified and exhibits low computational cost. Finally, the simulation results show that the proposed control methodology can provide robust tracking of velocity and altitude reference trajectories in the presence of system uncertainties.