State estimation of nonlinear time-varying complex networks with time-varying sensor delay for unknown noise distributions

The state estimation for nonlinear time-varying complex networks is a common problem, especially for the systems with time-varying sensor delay. This paper presents an artificial immune strategy based on particle filtering algorithm to estimate the states. Firstly, considering the nonlinear time-varying complex networks with time-varying sensor delay, this paper gives a theoretical proof that particle filter can be used for the state estimation. On this basis, this paper adopts the artificial immune algorithm to solve the problem which is particle impoverishment in particle algorithm. This paper gives the convergence of the proposed algorithm theoretically. Furthermore, for nonlinear time-varying complex networks with unknown noise distributions, this paper gives an improved algorithm which uses an improved kernel density estimation algorithm which has window width varying with probability density to estimate noise distributions. Simulation results demonstrate the effectiveness of the scheme, especially when nonlinear time-varying complex networks have time-varying sensor delay and noise distributions are unknown.