Adaptive Consensus Control of High-Order Uncertain Nonlinear Multi-agent Systems with Fuzzy Dead-Zone
Maryam Shahriari-kahkeshi,Abed Afrush,Viet-Thanh Pham
DOI: https://doi.org/10.1007/s40815-020-01005-6
IF: 4.085
2021-02-05
International Journal of Fuzzy Systems
Abstract:This paper considers the distributed consensus control of <i>n</i>th-order uncertain nonlinear multi-agent systems (MASs) with fuzzy dead-zone (DZ). Unlike the existing works which describes the DZ constraint by the conventional deterministic models, this work describes the DZ by the fuzzy and uncertain value. Therefore, some imposed restricting assumptions on the DZ parameters such as known bounds of the DZ slopes, or linear and deterministic model for DZ description which exist in the previous works are removed in this work. In the proposed scheme, the neural network (NN) is invoked to model the uncertain nonlinearities of the follower agents and uncertain interaction of their neighbors. To decrease the number of adaptive parameters and computational load, maximum norm of the weight coefficients of the NN is considered as an adjustable parameter and tuned online. Also, to decrease the controller complexity and eliminate the repeated derivatives of nonlinear functions in the controller, the dynamic surface control (DSC) method is used to develop the proposed distributed consensus scheme. Stability analysis based on the Lyapunov direct method guarantees that all of the signals in the closed-loop MAS are semi-globally uniformly ultimately bounded and the consensus error converges to a bound dependent on the design parameters. Therefore, appropriate choice of the design parameters makes the bound small. Simulation results on a well-known example demonstrate the effectiveness of the developed consensus algorithm.
computer science, information systems,automation & control systems, artificial intelligence