[Thesis] Limited time consistency and collision avoidance control algorithm of multiple underwater vehicles
Summary
This paper studies the finite time position consistency and collision avoidance problems of multiple underwater vehicle systems. First, based on the homogeneous control method, for leaderless and leader-follower multi-underwater robot systems, a finite-time position consistency algorithm that does not consider the collision between underwater robots is proposed. Specifically, in the case of leader-follower, a new distributed finite-time observer was developed to allow followers to estimate the leader's velocity. Secondly, by constructing the collision avoidance function and the connectivity maintenance function, an improved consensus algorithm containing the corresponding gradient terms for the multiple underwater vehicle systems in the two cases is proposed to ensure collision avoidance, connectivity maintenance, speed matching, and uniform bounds. . The simulation results show the effectiveness of the proposed control algorithm.
1 Introduction
Distributed collaborative control of multiple underwater robots has attracted much attention in recent years. This is not only due to the wide range of applications of underwater robots, such as ocean surveys, ocean exploration and deep-sea inspections (Fossen, 2002 and cited therein), but also due to the advantages of some multi-AUV systems compared to individual AUV systems, such as higher Higher efficiency, better robustness, and larger service areas (Cui, General Electric, & Choo, 2010; Hou, Xie Xiaohua, 2011; Yang & Gu, 2007). The distributed coordinated control of multiple underwater robot systems requires that each underwater robot uses only limited local information to achieve the overall goal in a complex underwater environment, which brings great challenges to the control. In order to improve the collaborative performance of multiple underwater vehicle systems, various methods have recently been studied, such as back-off technology (Cui et al., 2010; Yang & Gu, 2007) and adaptive control methods (Hou & Cheah, 2011), to name just a few A few examples.
Convergence speed is an important issue in the field of distributed cooperative control. Most of the existing distributed control algorithms for multiple underwater vehicles have asymptotic convergence, that is, the convergence rate of the closed-loop system is exponential at most when the stability time is infinite. In other words, consensus and formation cannot achieve all expected cooperative behaviors within a limited time. Aware of this fact, a finite time distributed control algorithm is preferable. In addition to faster convergence speed, closed-loop systems with finite time convergence usually show other good characteristics, such as better interference suppression characteristics and better robustness to uncertainties (Bhat & Bernstein, 2000) . In addition to faster convergence speed, closed-loop systems with finite time convergence usually show other good characteristics, such as better interference suppression characteristics and better robustness to uncertainties (Bhat & Bernstein, 2000) . Based on the above advantages and the core role of the consensus problem, several first-order finite-time consensus algorithms (Chen, Lewis, & Xie, 2011; Cortes, 2006) and second-order multi-agent systems (Khoo, Xie, & Man, 2009) have been developed recently. ; Wang Hong, 2008). In the problem of distributed limited time control, our group has also achieved some results, such as consensus algorithm (Du, Li, & Ding, 2013; Du, Li, Shi, 2012; Li, Du, Lin, 2011), attitude synchronization space Spacecraft (Du, Li, & Qian, 2011), and formation control of multiple incomplete mobile robots (Ou, Du, & Li, 2012).
In fact, when agents work together, conflicts may occur between them, which is very undesirable. Therefore, collision avoidance has always been another important issue for the success of collaborative control. Generally, collision avoidance methods can be divided into behavior-based methods (Wang, Yadav, & Balakrishnan, 2007) and potential field methods (Hokayem, Stipanovic, & Spong, 2010); Hou, Xie Xiaohua, 2011; Meng, Lin, & Ren , 2012; OlfatiSaber, 2006). Olfatil-Saber (2006) pointed out that in addition to avoiding collisions, connection maintenance and speed matching are also two equally basic rules for the coordination of multi-agent systems in practice. Specifically, connection maintenance refers to maintaining a certain form of connection between agents, which is usually done by using latent field methods (Bullo, Cortes, & Martinez, 2009; Meng et al., 2012; Olfati-Saber, 2006).
to be continued
6 Conclusion
In this paper, for the leader-less and leader-follower multi-underwater robot systems, the position consensus algorithm without/with collision is studied. Without considering the collision of the underwater robot, the control algorithm based on the homogeneous control method realizes the finite time position consistency of the underwater robot. Considering collisions, the modified position consistency algorithm including the gradient term of the potential function can not only guarantee collision avoidance, connectivity maintenance and speed matching, but also guarantee the consistency of bounds, that is, auv asymptotically converges to a bounded static or dynamic area.