Designing Iterative Learning Schemes for Cooperative-Antagonistic Systems With Random Access Communication Protocols
Wenjun Xiong, Xiaoxiao Wang, Lei Zou, Guanrong Chen · IEEE Transactions on Neural Networks and Learning Systems · 2024
In this article, the design issue for an iterative learning controller is investigated for the cooperative-antagonistic system under the scheduling effects of random access protocol (RAP). In order to reflect the heterogeneous characteristic of the underlying system, the dynamics of each node in the cooperative-antagonistic system are described by a two-time-scale system. For the purpose of avoiding data collisions in signal transmissions, the so-called RAP is introduced to schedule the data exchanges among nodes, where the transmission opportunities of nodes are modeled by a sequence of random variables with certain transition probabilities. Considering that the mutual relationships may be cooperative and also competitive among nodes in many real-world networks, a novel cooperative-antagonistic-based iterative learning controller is developed to handle the tracking problem of the system dynamics. Sufficient conditions are obtained for the design of the controller parameters. Furthermore, the derived results are extended to the case that the transition probabilities for the RAP are partially unknown. Finally, a numerical example is presented to illustrate the effectiveness of the proposed iterative learning control (ILC) scheme.