Adaptive dynamic task allocation and robust control scheme for target tracking

Abstract

This paper proposes a dynamic task allocation and robust control scheme for the agents modelled as double-integrator systems. In the proposed approach, k-winners-take-all (k-WTA) strategy together with an adaptive finite-time dual neural network is used to realized the dynamic task allocation. The robust controller is developed based on disturbance compen-sation. The stability is proved in theory and numerical simulation illustrates the efficiency of the proposed scheme.