Multi-vehicle Coordinated Control Technology And Application Research Of Networked Autonomous Driving Based On Swarm Intelligence

In modern society,humans rely more and more on intelligent decision making in daily life.In the field of transportation,with the rapid development of wireless communication technology,the problem of traffic congestion can be solved by vehicles’collaborative driving,which is the ultimate goal of the development of the networked vehicle,and has attracted the attention of a large number of scholars.This article takes the networked autonomous driving vehicle as the research object,studies the collaborative driving problem of the vehicle with the swarm intelligent(SI)theory,and gives the strategy of multi-vehicle cooperative driving under different application scenarios.First,the network-linked platoon model and its communication architecture were introduced,and the structure of hierarchical cooperative driving system was clearly defined.Based on this,a linear control model of the networked autonomous platoon was designed.The model draws on the idea of consistency control,taking into account the driving state of the preceding vehicle and the leading vehicle,and through the stability analysis,the validity of the line control system was proved.Control.Subsequently,four typical application scenarios of multi-vehicle cooperative control strategy were designed:multi-vehicle platoon maintenance,multi-vehicle platoon switching,obstacle avoidance and self-adaptation,and several solutions were proposed by using the line control model.Because the line control model can only be used to complete the application of a specific scenario,this paper studied the theory of swarm intelligence(SI),in order to improve the robustness of cooperative control system,embodied the advantages of multi vehicle cooperative control by networking.The characteristics of Multi Intelligent Vehicle System(MIVS)was presented,a hierarchical structure was proposed,and the self-organizing collaborative mechanism in nature was also described.Finally,the flocking control algorithm was applied to the MIVS.The networked autonomous driving platoon linear control model was optimized.Aiming at four typical application scenarios of cooperative control strategy,the algorithm was designed specifically:in order to overcome the"split"phenomenon,the concept of virtual guided vehicle was introduced;in order to solve the rigid structure problem of multi-intelligence entanglement,the perturbation force was introduced;a virtual agent was used to enhance the driving comfort;and the adaptive strategy of the vehicle encountered obstacle was discussed.In order to verify the effectiveness of networked driving linear control model and SI control algorithm,the flocking control algorithms were firstly simulated on the Simulink-CarSim co-simulation platform.The platform considered vehicle kinematics and dynamics characteristics.The simulation implemented the SI control algorithm that applied in the multi-car collaborative driving scenario.Simulation results show that the algorithm was effective.Finally,on the Dongfeng AX7 networked autonomous driving vehicle platform,this paper validates the practicality of multi-vehicle cooperative control algorithm based on SI control algorithm through real vehicle testing.In a north China vehicle proving ground,a platoon of three AX7 had completed the test of start,acceleration\deceleration and parking at0~20km/h speed,keeping the distance in 15m(including body length)by multi-vehicle platoon maintenance.The control precision of steady state error is kept in 0.5m.Then the multi-vehicle platoon switching and obstacle avoidance were tested in a central China vehicle proving ground.The Shortest vehicle spacing in cooperative lane changing scene was controlled under 5m without body length.The strategy execution cycle was less than 20s.Maximum lateral acceleration was 2.3m/s~2.Lateral acceleration change rate was 0.1g/sec.The function of this control system-satisfied the demand of the design.Self-adaptation was tested in actual road with a complex environment.The static and dynamic obstacles were detected there.All above tests verified the effectiveness and application value of the multi-vehicle cooperative control strategy based on swarm intelligence in this paper.