Active Disturbance Rejection Predefined-Time Consensus Tracking for Nonlinear Multiagent Systems via Self-Triggered Communication

This paper addresses the active disturbance rejection predefined-time consensus tracking problem for non-affine nonlinear multiagent systems (MASs) with self-triggered communication. With the active disturbance rejection strategy, the uncertainty of the system can be estimated and the issue of explosion of complexity under the backstepping framework is handled. To improve the utilization of network resources, a self-triggered mechanism that the next trigger moment is calculated from the current information is adopted, for communication of each follower. The advantage of this triggered mechanism is that the triggering conditions are not required to be continuously monitored. Meanwhile, a time-varying tuning function-based predefined-time method is designed to achieve convergence within the predetermined time. Compared with existing finite/fixed-time convergence consensus schemes, its biggest advantage is that the upper bound of convergence time can be a positive constant arbitrarily selected by users. Additionally, the problem of input saturation is considered in the controller design to improve the generality of our scheme. Finally, the effectiveness of the proposed scheme is verified via simulation example. Note to Practitioners—In this paper, the active disturbance rejection predefined-time consensus tracking control problem for MASs with non-affine properties is discussed, which is embodied in numerous applications, such as sensor networks, space satellites and robot formations. The convergence time required for all agents to conclude consensus is used as an important indicator to evaluate the control performance in the consensus control of MASs. The existing control methods including finite and fixed time methods can only achieve the system to converge in a preset time. Based on this, it is a challenging topic to achieve predefined-time convergence by setting the accurate convergence time t. Meanwhile, to improve the utilisation of network resources, a self-triggered mechanism is designed. The advantage of this triggered mechanism is that the triggering conditions are not required to be continuously monitored, which greatly reduces the control cost. Additionally, the problem of input saturation is considered in the controller design to improve the generality of our scheme.