The paper "Containment Control for Networked Unknown Lagrangian Systems with Multiple Dynamic Leaders under a Directed Graph" by Jie Mei, Wei Ren, Bing Li, and Guangfu Ma, has been accepted for presentation in 2013 ACC, to be held in Washington, DC, USA on June 17-June 19, 2013.
This paper studied the containment control with multiple dynamic leaders for networked Lagrangian systems under a general directed graph, in the presence of nonlinear uncertainties and external disturbances. Neural networks based method was used to approximate the unknown nonliearities. Adaptive gain design and sign functions were used in the proposed control algorithms. A amazing part was that, in the leaderless case, the consensus error converged to zero. The control algorithm without neighbors' velocity information was also proposed. The main result is proved by a two-step Lyapunov functions method. There are also some interesting questions remaining:
1) How to achieve zero-error tracking when the leader is verying and there exist unknown nonlinearities and bounded external disturbances?
2) How to achieve dynamic consensus (with constant but nonzero final velocity) with/without neighbors' velocity information? It seems that auxilary variables with integrtor is an effective way.
This paper studied the containment control with multiple dynamic leaders for networked Lagrangian systems under a general directed graph, in the presence of nonlinear uncertainties and external disturbances. Neural networks based method was used to approximate the unknown nonliearities. Adaptive gain design and sign functions were used in the proposed control algorithms. A amazing part was that, in the leaderless case, the consensus error converged to zero. The control algorithm without neighbors' velocity information was also proposed. The main result is proved by a two-step Lyapunov functions method. There are also some interesting questions remaining:
1) How to achieve zero-error tracking when the leader is verying and there exist unknown nonlinearities and bounded external disturbances?
2) How to achieve dynamic consensus (with constant but nonzero final velocity) with/without neighbors' velocity information? It seems that auxilary variables with integrtor is an effective way.