Abstract
This paper explores adaptive neural network output feedback control for a group of output-constrained systems using dynamic cooperative learning. The barrier Lyapunov function is employed to constrain the outputs within the required range. For all agents, we use radial basis function neural networks to identify their unknown dynamics. The undirected and connected communication topology is used for exchanging learned knowledge of each agent. Therefore, the approximation domain of learned neural networks is extended. Finally, the stored constant neural weights are used as previous experiences to construct new neural controllers, which not only can be used for the same control tasks, but also can improve dynamic performance and reduce the computational burden. Finally, the approach feasibility is demonstrated by simulation results.
Funding Information
  • National Natural Science Foundation of China (62163030)
  • Natural Science Foundation of Inner Mongolia of China (2020BS06004)
  • Research Foundation for Advanced Talents Research Foundation of Inner Mongolia University (21700-5185130)

This publication has 45 references indexed in Scilit: