Prescribed finite-time adaptive tracking control for a class of full state constrained non-strict feedback nonlinear multi-agent systems

Abstract The target of this paper is to discuss the prescribed finite-time tracking control issue for uncertain full state constrained non-strict feedback nonlinear multi-agent systems (MASs). An improved distributed adaptive control protocol is devised builded on a developed universal Lyapunov barrier function (BLF) approach and radial basis function (RBF) neural networks (NNs) technique. Compared with previous results, the proposed control strategy has the following characteristics. Firstly, under the backstepping design framework, the designed practice virtual control signals can satisfy the same limitation as the corresponding state variables so that the MAS is able to complete the tracking task smoothly. Furthermore, the proposed virtual and real control signals do not contain error terms whose fractional power is less than 1, then the singularity problem is avoided successfully. Secondly, with pre-specified settling time and accuracy level, the tracking error can converge to a small enough neighbourhood of origin without violating the constraints on state variables during the operation. It is shown that the proposed protocol enables the MAS achieve synchronous tracking within a pre-given finite time, meanwhile, signals in the MAS are always bounded. At last, the availability of the proposed control protocol is verified by a numerical simulation example.