Enhanced Recurrent Fuzzy Neural Fault-Tolerant Synchronization Tracking Control of Multiple Unmanned Airships via Fractional Calculus and Fixed-Time Prescribed Performance Function

This article proposes a fractional-order intelligent fault-tolerant synchronization tracking control (FO-I-FTSTC) scheme for multiple unmanned airships (UAs) against actuator faults. Within the developed control architecture, fixed-time prescribed performance functions (PPFs) are first designed to transform the synchronization tracking errors into a new set of error variables, such that the original errors are strictly confined within the prescribed bounds. Then, fractional calculus and sliding mode surface are sequentially introduced to construct the FO errors. Moreover, to handle the unknown terms and bias faults in the FO sliding-mode error dynamics, fuzzy neural networks with recurrent loops are artfully constructed to act as the intelligent learning units. Furthermore, the norm of the loss-of-effectiveness fault factors is introduced for each UA to reduce the number of adaptive parameters. The distinct feature of the proposed method is that the FO-I-FTSTC performance is significantly enhanced by integrating recurrent fuzzy neural networks, fractional calculus, and fixed-time PPFs into a unified framework, leading to a high-precision control scheme. It is shown by Lyapunov analysis that all UAs can track their desired references in a synchronized manner, and the synchronization tracking errors are bounded and strictly confined within the prescribed error bounds. Comparative hardware-in-the-loop experiments are presented to show the effectiveness of the proposed FO-I-FTSTC scheme.

I. Introduction

Recently, various unmanned aerial systems have been employed to provide high efficiency and reduce the human operation risk in many fields, such as forest fire monitoring [1], [2], smart city observation [3], and search and rescue [4]. Among unmanned aerial systems, unmanned airships (UAs) can provide durative observation due to the long-term hovering feature [5]. More recently, numerous cooperative control strategies have been proposed for multiple agents with the single/double integral or linear forms to achieve the regulation, trajectory tracking, or containment control objectives [6]–[8]. Arising from the cooperative control of multiple agents, the cooperation of multiple UAs can significantly increase the efficiency [9]. However, the existing cooperative control methods developed for multiple agents may not be directly extended to cooperatively control the multiple UAs since the strong nonlinearities are frequently involved into the UA dynamics [10]. Moreover, unexpected actuator faults are rarely considered in the cooperative control design for multiple agents with the single/double integral or linear forms, which may further limit the implementation of existing cooperative control strategies in the multiple UAs.