Formation control and coordination of multiple unmanned ground vehicles in normal and faulty situations: A review

Abstract

Recently, multiple unmanned vehicles have attracted a great deal of attention as viable solutions to a wide variety of civilian and military applications. Among many topics in the field of multiple unmanned systems, formation control and coordination is of great importance. This paper presents a comprehensive literature review on the strategies and methodologies applied for formation control of multiple unmanned ground vehicles in both normal and faulty situations. First, the basic definitions of formation control and coordination are provided as well as their classification. Second, a comprehensive literature review of formation control strategies is introduced. Moreover, an overview on fault detection and diagnosis and fault-tolerant cooperative control of UGVs is presented. Finally, open problems, challenges, and future directions are highlighted.

Introduction

The rapid development of advanced mechatronic, computing and communication technologies in the last two decades enables a new robotic system that can interact with both human and other robots in a cooperative manner. This technology is called network robot systems (NRS) (Sanfeliu, Hagita, & Saffiotti, 2008) or multiple unmanned vehicles (MUVs) (Zhang & Mehrjerdi, 2013). As compared to a single vehicle, the usage of (MUVs) has many advantages such as:

• Multi-tasking: When using a team of robots, the task can be decomposed into several sub-tasks which can be handled simultaneously. Therefore, the mission can be achieved much faster than a single robot, resulting in time reduction of mission execution. For example, using a team of MUVs in forest monitoring and fire detection mission can reduce the time of monitoring and information collection significantly as compared with a single UAV;

• Fault-tolerance: If fault occurs in one or more robots in the team, the other robots can be appropriately reconfigured to mitigate the negative effect on mission execution. This essentially increases the overall system robustness and reliability, especially in dangerous missions;

• Cost-effectiveness: Designing a powerful and versatile robot that is capable of handling different tasks might not be feasible due to limits of robot size and payload. However, using a group of robots with various functionalities, cost-effective robots can be built without losing the capability of different tasks handling;

• Flexibility: The functionality of a group of robots can be easily reconfigured by combining different robots with different payloads; and

• Distribution: Robots can work simultaneously at different positions in the same workspace. For example, during a surveillance task, the target can be monitored from different positions with different types of sensors. Consequently, more detailed and accurate information about the target can be obtained.

Due to these advantages, robotic networks are applied in both military and civil applications including surveillance (Acevedo, Arrue, Maza, Ollero, 2014, Kingston, Beard, Holt, 2008, Mesbahi, Hadaegh, 2001), search and exploration (Franchi, Freda, Oriolo, Vendittelli, 2009, Howard, Parker, Sukhatme, 2006, Hu, Xu, Xie, 2013, Nieto-Granda, Rogers, Christensen, 2014), cooperative reconnaissance (Balch & Arkin, 1998), environmental monitoring (Dunbabin, Marques, 2012, Marques, Martins, de Almeida, 2005), and cooperative manipulation (Prasad, Sharma, Vanualailai, 2016, Tanner, Loizou, Kyriakopoulos, 2003), respectively. During mission execution, vehicles are required to travel autonomously between different locations, to avoid collision of obstacles and team members, and to accommodate faults in individual members, respectively.

The main topics in the field of multi-robot systems are (Arai, Pagello, & Parker, 2002): i) biological inspirations; ii) communication; iii) architectures, task allocation, and control; iv) localization, mapping, and exploration; v) object transport and manipulation; vi) motion coordination; and vii) formation reconfiguration.

Within these topics, formation control and coordination receive a great deal of attention because a group of robots can accomplish a mission more effectively by maintaining a pre-defined formation shape. Formation control is inspired by the emergent self-organization observed in nature, like birds flocking and fish schooling (Xie, 2007). Formation control has been extensively studied in the literature, with application to the unmanned ground vehicles (UGVs) (Ailon, Zohar, 2012, Consolini, Morbidi, Prattichizzo, Tosques, 2008, Do, 2008, Liang, Liu, Wang, Chen, Xing, Liu, 2016, Liu, Tan, Liu, 2007, Liu, Tan, Liu, 2007, Mehrjerdi, Saad, Ghommam, 2011, Monteiro, Bicho, 2010, Sira-Ramírez, Castro-Linares, 2010, Xiao, Li, Chen, 2016), unmanned aerial vehicles (UAVs) (Abdessameud, Tayebi, 2011, Anderson, Fidan, Yu, Walle, 2008, Dong, Yu, Shi, Zhong, 2015, Dong, Zhou, Ren, Zhong, 2016, Karimoddini, Lin, Chen, Lee, 2011, Qiu, Duan, 2014, Sen, Sahoo, Kothari, 2017, Wang, Xin, 2013, Zhang, Duan, Yu, 2010), and autonomous underwater vehicles (AUVs) (Cui, Ge, How, Choo, 2010, Das, Subudhi, Pati, 2016, Li, Xie, Yan, 2016, Panagou, Kyriakopoulos, 2013, Qi, Cai, 2016, Wang, Yan, Li, 2012), respectively.

The formation control problems can be mainly summarized as: i) formation shape generation; ii) formation tracking; iii) formation reconfiguration and selection; and iv) task assignment in formation.

With respect to the aforementioned issues, there are two major challenges. The first one lies in selection of the formation control strategy able to: i) achieve the desired mission; ii) cover the main formation control problems (if exist in the desired mission); iii) maintain the formation stability; and iv) be implemented in real-time. The second challenge is to design a formation controller that is computationally simple, robust, reliable, fault-tolerant, and can be implemented in real time.

In addition to maintaining a formation shape during task executions under normal conditions, it is highly desirable that robots possess a fault-tolerance capability against faults. Thus, the healthy robots can react correspondingly to eliminate the adverse effect on mission completion. Otherwise, the formation shape will be broken, leading to failed missions. Such an objective can be achieved by so-called fault-tolerant cooperative control (FTCC) or fault-tolerant formation control (FTFC) strategies (Kamel, Yu, & Zhang, 2018). FTFC is considered as a special case of FTCC where it is required for the team members to maintain a specific formation during the mission execution. Accordingly, integrating FTCC/FTFC algorithm to a team of MUVs becomes very important from the aspects of safety, reliability, and mission completion. The main challenges here are: i) how to detect the faults, estimate their amplitudes, and determine degree of severity; ii) how to take the decision to compensate the fault effect on the mission completion; and iii) how to execute this decision.

This paper is intended to review and highlight the existing research of formation control and coordination of multiple UGVs in normal cases where no fault occurs for any vehicle in formation, and faulty cases where one or more vehicles subject to faults. This review can be divided into two main sections: i) an overview of formation control and coordination, its basic definitions, classifications, and strategies in a fault-free case; and ii) a review of formation control under faulty situations in a particular effort on fault detection and diagnosis (FDD) and FTCC algorithms.

The rest of this paper is organized as follows. Section 2 presents a comprehensive review of formation control and coordination in fault-free cases. In Section 3, a review of FTCC and FDD applied to UGVs in highlighted. Challenging issues and main future directions are presented in Section 4. Finally, concluding remarks are drawn in Section 5.