Scholarlink

<p>The development of autonomous driving technology requires extensive validation, but real-world testing presents significant cost and feasibility challenges. CARLA, an open-source simulator based on Unreal Engine, offers a high-fidelity virtual environment with sensor simulation and ROS support. However, the default PhysX vehicle model in CARLA lacks consistency in dynamic behavior, leading to a Sim-to-Real Gap. In contrast, vehicle dynamics simulators such as TruckSim provide high-accuracy motion predictions but offer limited sensor compatibility and rendering capabilities. To enhance the realism of vehicle dynamics while maintaining CARLA’s comprehensive simulation environment, this study integrates TruckSim with CARLA. The consistency of vehicle behavior is evaluated by comparing standalone TruckSim with the CARLA-TruckSim integrated model under controlled conditions. Using CARLA and TruckSim, vehicle responses to various driving scenarios are analyzed to ensure dynamic accuracy. Additionally, the feasibility of applying high-fidelity vehicle models to CARLA is examined, particularly in complex terrains such as off-road and uneven surfaces, to support advanced autonomous driving research.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Integration and Validation of Trucksim Vehicle Model on Carla Simulator

  • 期刊论文

  • 发表日期：

    2026年2月20日

查看全文
  • Tae Ho Ji,
  • Yong Jin Kim,
  • Jun Young Jung,
  • Chibum Lee

摘要

The development of autonomous driving technology requires extensive validation, but real-world testing presents significant cost and feasibility challenges. CARLA, an open-source simulator based on Unreal Engine, offers a high-fidelity virtual environment with sensor simulation and ROS support. However, the default PhysX vehicle model in CARLA lacks consistency in dynamic behavior, leading to a Sim-to-Real Gap. In contrast, vehicle dynamics simulators such as TruckSim provide high-accuracy motion predictions but offer limited sensor compatibility and rendering capabilities. To enhance the realism of vehicle dynamics while maintaining CARLA’s comprehensive simulation environment, this study integrates TruckSim with CARLA. The consistency of vehicle behavior is evaluated by comparing standalone TruckSim with the CARLA-TruckSim integrated model under controlled conditions. Using CARLA and TruckSim, vehicle responses to various driving scenarios are analyzed to ensure dynamic accuracy. Additionally, the feasibility of applying high-fidelity vehicle models to CARLA is examined, particularly in complex terrains such as off-road and uneven surfaces, to support advanced autonomous driving research.