Practical Fixed-Time Multi-Group Time-Varying Formation for Second-Order Multi-Agent Systems with Actuator Attacks and Collision Avoidance

Xisheng Zhan,Rongxiang Lu,Jie Wu,Huaicheng Yan
DOI: https://doi.org/10.1016/j.ins.2024.120821
IF: 8.1
2024-01-01
Information Sciences
Abstract:In this paper, the fixed-time multi-group time-varying formation problem for second-order nonlinear multi-agent systems (MASs) is studied. It takes into account some factors such as actuator attacks, collision avoidance and Markov switching topology. Based on the presence of actuator attacks and other unknown nonlinearities, radial basis function neural network (RBFNN) is used to approximate these unknown functions. Following this, a distributed fixed-time RBF sliding manifold is proposed to enable the system to reach the desired state in a fixed time. Additionally, a collision-avoiding repulsive potential function is introduced to prevent agents from colliding with each other during the formation process. By combining these factors, a control strategy is proposed that can cope with switching topologies and enable all agents to achieve practical fixed-time multi-group formations. Finally, two examples are provided to demonstrate the feasibility of the control protocol.
What problem does this paper attempt to address?