Abstract: The traditional research of vehicle autonomous lane changing is based on the ideal road curvature and simplified vehicle motion, the vehicle path data are directly transmitted to the controller by matching the sampling period, which is difficult to verify the reliability of the system in the real environment. This paper designs a multi-sensor layout scheme and fusion strategy, and constructs the highway simulation scene on the Matlab/Simulink platform to verify the effectiveness of the multi- sensor fusion strategy. The simulation results show that the multi-sensor fusion strategy can realize the continuous tracking of the target vehicle in the high-speed scene, simulate the environmental perception and data fusion process of intelligent vehicles, and then provide technical reference for the development of intelligent vehicle autonomous lane changing system.

Key words: Intelligent vehicle; Highway; Autonomous lane change; Multisensor fusion strategy; Joint probabilistic data association (JPDA)

摘要： 传统车辆自主换道研究多以理想道路曲率及车辆间简化的相对运动为前提，通过匹配 采样周期将车路数据直接传入控制器，难以验证系统在真实环境下的可靠性。文章设计了一 种多传感器布置方案及融合策略，并在 Matlab/Simulink 平台构建高速公路仿真场景，验证多 传感器融合策略的有效性。其仿真结果表明，该多传感器融合策略能够在高速场景下实现对 目标车辆的持续跟踪，模拟智能车辆环境感知、数据融合过程，进而为智能车辆自主换道系 统的开发提供技术参考。

关键词: 智能车辆；高速公路；自主换道；多传感器融合策略；联合概率数据关联（JPDA）