Consensus Control for Directed Networks Under Quantized Information Exchange

Abstract

In this paper, without assuming balanced network topologies, we address the weighted average consensus problem for discrete‐time single‐integrator multi‐agent systems with logarithmic quantized information communication. By incorporating generalized quadratic Lyapunov function with the discrete‐time Bellman–Gronwall inequality, a new upper bound about the quantization precision parameter of the infinite‐level logarithmic quantizer is derived to design quantized protocol, under which agents in strongly connected directed networks can attain weighted average consensus. The obtained new upper bound clearly characterizes the intimate relation between the quantization precision parameter and the directed network topology. The proposed quantized protocol is particularly applicable to digital networks where balanced message passing among agents is not available.

Distributed Consensus Filter for a Class of Continuous‐Time Nonlinear Stochastic Systems in Sensor Networks

“…Þdω t ð Þ: (7) x(t) is said to be exponentially ultimately bounded in mean square if there exist positive constants c 1 , c 2 , c 3 such that for all t ≥ 0,…”

“…Lemma 1 [30]. Consider the stochastic differential (7). Assume that there exists a function V(x, t) ∈ C 2 , 1 and positive constants c 1 , k 1 , k 2 such that…”

“…Compared with the previous works , the contribution of this paper is at least fivefold. First, in contrast to the works of , which assume that the target has linear dynamics or target states are available, this paper considers that the target has nonlinear dynamics and its states are not available and a DCF is proposed to estimate the states of the target.…”

“…First, in contrast to the works of , which assume that the target has linear dynamics or target states are available, this paper considers that the target has nonlinear dynamics and its states are not available and a DCF is proposed to estimate the states of the target. Second, a DCF is proposed in this paper that can be used for continuous nonlinear systems, whereas most of the previous results in are discrete and cannot be applied for state estimation in continuous‐time systems. Third, in contrast to the work of , in which stability analysis is limited to linear systems, in this paper stability analysis is carried out in a nonlinear system and ultimate boundedness of filter is proved in the sense of mean square.…”

“…These types of algorithms are called distributed consensus algorithms. Recently, many distributed consensus algorithms have been introduced and applied [6][7][8][9][10][11][12][13][14][15]. These algorithms can be divided into three groups.…”