摘要
We study the consensus problem in a weighted and directed network composed of self-propelled agents. The weight which quantifies the relationship between the agents is adjusted according to their state incoherence to suppress the heterogeneity in state. The proposed consensus protocol enhances the convergence efficiency of consensus greatly and has better performance than that of two other protocols. The convergence efficiency can be further improved by adjusting the parameter of this protocol. Furthermore, the robustness of the system against coupling time-delay is significantly increased. Unlike most consensus acceleration methods, no topological information is needed in our model. Networks with different stuctures are investigated to demonstrate the generality of the method.