基于多智能体系统车辆跟驰模型的数值模拟

Numerical Simulation of Car Following Model Based on Multi-Agent System

摘要: 近年来，交通拥堵问题已经越来越严重并且受到国内外学者的广泛关注。本文主要研究基于多智能体系统的车辆跟驰模型，首先阐述了多智能体系统与车辆跟驰模型，主要考虑所有车辆单车道行驶且每辆车动力学方程可以表示为简单的二阶线性微分方程。其次，根据多智能体系统协同控制思想，为跟驰模型中每辆车设计新的控制输入，最终使得所有车辆速度一致，不发生碰撞且相邻车辆之间距离保持不变，即所有车辆达到跟驰状态。最后，当头车匀速和匀减速行驶时，分别给出相应的数值模拟结果验证本文所提出的跟驰模型的有效性。

Abstract: In recent years, the traffic congestion problem has become more and more serious and has been widely concerned by scholars at home and abroad. Firstly, this paper describes the multi-agent system and the car following model which mainly considers that all vehicles travel on a single lane and the dynamic equation of each vehicle can be expressed as a simple second-order linear differ-ential equation. Secondly, based on the cooperative control idea of multi-agent system, a new con-trol input is designed for each vehicle in the car following model, so that all vehicles have the same speed, no collision and the distance between adjacent vehicles remains unchanged, that is, all vehi-cles reach the car following state. Finally, when the leading vehicle moves at a constant speed and a uniform deceleration, the numerical simulation results are given respectively to verify the effec-tiveness of the car following model proposed in this paper.