Opinion Dynamics Based On DeGroot Model And Its Application In Distributed Target Recognition

In recent years,with the rapid development of science and technology,many important changes have taken place in the communication forms and modes of public opinions.It is characterized by immediacy,freedom and large amount of information.Therefore,a sound public opinion guidance mechanism has become a common concern of governments all over the world.French and DeGroot first proposed an agent-based public opinion formation model,that is,each agent updates his own opinion by communicating with his neighbors.In this paper,the further application of DeGroot model in public opinion dynamics and information fusion is studied,which is respectively the public opinion dynamics model of stubborn agent with sensitive function and the multi-sensor target recognition information fusion method based on multi-ground parallel fusion:(1)We study a dynamic model of opinion formation,in which the susceptibility of each agent to be influenced depends on his current opinion.This notion is represented by a susceptibility function.Specifically,we propose two different susceptibility functions to describe two classes of agents who are called the stubborn positives and the stubborn neutrals,respectively.We suppose that the opinion values are scalars on a continuous interval.We supply theoretical analysis to our specific susceptibility model and develop necessary and sufficient conditions for the final consensus value,and provide simulations to illustrate these results.(2)This paper provides a kind of multi-points parallel information fusion for target recognition of multi-sensors,and each sensor only exchanges information with their neighbors.Based on Dempster–Shafer evidence theory,this paper designs a distributed fusion algorithm,by which each sensor can establish consensus target recognition result under some topologies.Finally,we use simulation to verify the validity of this distributed fusion algorithm.