Adaptive finite-time tracking control for non-cooperative targets in multi-agent systems via active replacement...

This paper proposes a finite-time tracking control strategy for multi-agent systems executing non-cooperative target-tracking missions within a designated area. A distributed controller is developed that integrates a radial basis function (RBF) neural network to approximate the unknown nonlinear dynamics of both the target and agents. To accelerate tracking convergence and enhance overall efficiency, an active replacement mechanism is introduced. This strategy proactively substitutes agents exhibiting the largest tracking errors with better-performing candidates. Additionally, an adaptive controller with time-varying gains is developed to explicitly avoid actuator saturation. Theoretical analysis demonstrates that the proposed approach achieves finite-time convergence, and simulations validate its effectiveness in reducing convergence time through active agent replacement.