Adaptive Dynamic Planning Consensus Control of Multi Agent Systems Based on Stochastic Fuzzy Neural Networks

This paper proposes an SFNN-ADHDP algorithm to approximate the solution of the highly coupled HJB equation for the optimal consensus control problem of first-order nonlinear multi-agent systems. Optimal consensus control of multi-agent systems is achieved by innovatively combining stochastic fuzzy neural networks with adaptive dynamic programming. The proposed SFNN-ADHDP algorithm only requires the topology information of multi-agent systems for optimal consensus control. Graph theory related knowledge and consensus problem modeling for the optimal consensus control problem of multi-agent systems are firstly presented in this paper. Secondly, the specific algorithm structure and implementation details of the SFNNADHDP are proposed. Finally, a set of simulation experiments are conducted to demonstrate the excellent anti-interference ability and superiority of the proposed algorithm, comparing the SFNN network with the ANFIS network by attempting to incorporate Gaussian white noise and other methods into the network.