Formation‐containment control of networked collision‐free Lagrangian systems

Abstract

Summary The distributed formation‐containment (DFC) problem under a directed graph is addressed for networked Euler‐Lagrange systems. First, using a leader‐follower framework, the DFC problem is properly defined. For the leaders and the followers, respectively, a DFC control law is next proposed without using velocity information. Based on the artificial potential function, all the agents can achieve the control objective satisfactorily while avoiding collisions with others as well as the obstacles in the environment. By the Lyapunov stability theory, the boundedness of the error signals is guaranteed. Simulations are finally given to show the feasibility of this approach.

Containment control of discrete‐time multi‐agent systems with application to escort control of multiple vehicles

“…Figures 13,14,and 17 of Example 4 are consistent with those figures of Example 1, which indicate that the convergence of tracking errors and that the escorted following vehicles move into the convex hull formed by scout leading vehicles. Besides, Figures15 and 16show the convergence of velocity and acceleration accordingly.…”

“…Note that state-of-the-technology partial driving automation provides a practical basis to perform the escort control on freeways since both the longitudinal and lateral movements can be performed using the local information perceived by vehicles. Evidently, the escort control problem can be formulated into a containment control problem, [16][17][18][19][20][21] called "Loyal scout systems." The scout leaders perform surveillance, reconnaissance, and early warning missions while encircling a safe area.…”

“…The agents usually can be categorized as leaders and followers, where leaders and followers are unidirectional and bidirectional coupled dynamic systems, respectively. Generally speaking, according to the number of leaders in the network cooperative control problem of MAS can be mainly classified into three categories including the leaderless consensus problem (zero leader), 3‐10 the leader‐follower tracking problem (single leader), 11‐15 and the containment control problem (multiple leaders) 16‐21 …”

“…This problem is known as the containment control problem, which admits various practical applications, such as a brand‐new military project “Loyal Wingman” system * . Li et al 17 proposed a control law without using velocity information to solve the distributed formation‐containment problem while avoiding collisions. Cao et al 16 studied distributed containment control for double‐integrator dynamics in the presence of both stationary and dynamic leaders.…”

“…the leaderless consensus problem (zero leader), [3][4][5][6][7][8][9][10] the leader-follower tracking problem (single leader), [11][12][13][14][15] and the containment control problem (multiple leaders). [16][17][18][19][20][21] The leaderless consensus problem aims to drive state/output of agents to achieve consensus by choosing an appropriate distributed controller. In practice, apart from forming the consensus of state/output configuration, it is also desired that the whole distributed systems can track the trajectory of the (single) leader, that is, the state/output of followers follows that of the leader, which forms the leader-follower formation tracking problem.…”