This paper concerns partial topology identification problem of stochastic delayed complex dynamical network. Compared with many network models, the novel output coupling is introduced to the network model of this paper, which does not require all state variables of the vertex to be observable. By utilizing the graph-theoretic method and adaptive pinning control technology, the identification criteria of multi-weighted stochastic complex dynamical networks are obtained. Ultimately, the Lorenz system and small-world network are chosen to do numerical simulations. The simulation results indicate that the successful identification time (SIT) of stochastic delayed complex dynamical network with output coupling is reduced by about 56% on average by comparing with the SIT of the network without output coupling, which implies the superiority of output coupling.
Read full abstract