Observer-Based Adaptive Neural Networks Bipartite Containment Control of Multiagent Systems With Input Quantization

Abstract:
This paper studies the bipartite containment control problem on directed graphs for consistent quantization of nonlinear multi-agent systems (MASs) with unmeasurable states. The neural network observer and neural network logic system are used to estimate the unknown state and approximate unknown nonlinear function respectively. By designing a suitable distributed protocol, the followers are programmed to converge into a convex hull comprising the trajectory of each leader and the opposite trajectory of different signs. By combining the adaptive backstepping technology and the first-order filtered signal, for each follower, an observer-based neural network adaptive quantization control mechanism is proposed. It is proved that the semi-global uniform ultimate boundedness of all signals in a closed-loop system can be ensured. Finally, a simulation model illustrates the effectiveness of the control method.