基于模型预测控制算法的车辆编队研究  

Research on Vehicle Platoon Based on Model Predictive Control Algorithm

作　　者：李启朗 郭文静 LI Qilang;GUO Wenjing(School of Mathematics and Physics,Anhui Jianzhu University,Hefei 230601,China)

机构地区：[1]安徽建筑大学数理学院,安徽合肥230601

出　　处：《安徽建筑大学学报》2024年第5期40-45,共6页Journal of Anhui Jianzhu University

基　　金：安徽省高校省级自然科学研究项目(2022AH050252)。

摘　　要：为应对自动驾驶车辆运动的模型不确定性问题,根据协同自适应巡航控制(CACC)框架,搭建车辆间纵向运动学模型,并建立相应的离散状态空间方程。采用模型预测控制(MPC)方法,预测前车或车队未来状态,以优化跟随车辆的运动控制。通过车间通信获取跟随车辆与前车的信息,从而得出跟随车辆的期望加速度。通过对五辆车的编队仿真实验,验证了所提编队控制器的有效性。该方法以加速度作为控制量,更符合实际应用场景。与PID算法对比,该方法的车辆最大加速度降低了19.23%,最大位置跟随误差降低了63.51%。对不同车头时距的情况进行了仿真研究,结果表明车头时距越小,车辆跟踪效果越佳。To address the uncertainty in the dynamics of autonomous vehicle model,a longitudinal kinematic model for vehicle-tovehicle interactions is developed within the framework of cooperative adaptive cruise control(CACC),along with the corresponding discrete state-space equations.A model predictive control(MPC)approach is utilized to forecast the leading vehicle and the vehicle platoon,in order to optimize the motion control of the following vehicle.The vehicle-to-vehicle communication information is obtained to determine the desired acceleration for the following vehicle.The effectiveness of the proposed platoon controller is validated through simulation experiments involving a platoon of five vehicles.This method,which takes acceleration as the control variable,is better suited to practical applications.Compared to the PID algorithm,this method achieves a 19.23%reduction in maximum vehicle acceleration and a 63.51%reduction in maximum position tracking error.Simulation studies under different headway times indicate that shorter headway times result in improved vehicle tracking performance.

关 键 词：智能交通 协同自适应巡航 车辆编队控制 模型预测控制 

分 类 号：U491[交通运输工程—交通运输规划与管理]