Research On Dynamic Game Decision Among Multi-agents In Intelligent Networked Environment

Intelligent connected system has become a research hotspot in the field of intelligent transportation.Under the intelligent connected environment,mixed traffic flow composed of vehicles with different intelligence levels and connected levels is a necessary stage for the development of intelligent connected systems.The multi-agent composition of the intelligent networked system includes the intelligent networked cloud system and vehicles with different levels of intelligence and networked level in mixed traffic flow.The application of game theory to study the interaction between them is mainly divided into two aspects: on the one hand,from the perspective of vehicles,studying the game interaction relationship between vehicles in mixed traffic flows,analyzing game decisions and determining game returns,such studies have been relatively Mature,but this type of research is an extension of the traditional game driving decision-making game between vehicles,and it is not entirely appropriate to study mixed traffic flow in an intelligent networked environment.On the other hand,it is from the perspective of an intelligent networked cloud to study the game interaction between vehicles with different levels of intelligence and network connectivity in the intelligent networkconnected cloud mixed traffic flow.The research on game interaction in this aspect needs to be a dynamic cooperative game,while the research on dynamic cooperative games in the transportation field is relatively rare.This paper studies the dynamic game decision among multiple agents in the intelligent networked environment,establishes a dynamic cooperative game model and simulation platform among multiple agents in the intelligent networked environment,and provides better strategies for multiple agents in the intelligent networked environment based on the optimal cloud system.The main work of this paper is as follows:(1)This article analyzes the multi-agent composition and information interaction relationship in the intelligent networked environment,and correspondingly clarifies the mechanism mapping of the game theory,and specifically analyzes the game interaction mechanism between the multiple agents in mixed traffic flow in the intelligent networked environment.Then find the game and competitiveness relationship between them.These provide theoretical support for the establishment of the later dynamic cooperative game model.(2)This paper analyzes the multi-person alliance cooperation game model in mixed traffic flow,and establishes a dynamic cooperation game model among multiple agents in an intelligent network connection environment and solves the problem.From the perspective of the intelligent networked cloud,the model aims at the intelligent networked system optimization,and provides the optimal decision for vehicles in mixed traffic flow.At the same time,it encourages vehicles to adopt a cooperative behavior strategy that maximizes revenue,suppresses selfish behavior of selfish vehicles,and forms a stable and benign mixed traffic flow in the intelligent network.(3)The model parameters are calibrated in this paper,and their effectiveness is verified by simulation.By analyzing the simulation results of mixed traffic flow,we quantify the impact of the specific proportion of connected cars on the efficiency of traffic circulation.