A Digital Twin-Driven Simulation Platform for Autonomous Vehicle Testing and Algorithm Validation

Digital twin parallel systems have demonstrated significant advantages in improving testing efficiency and controlling research costs, and are widely regarded as a key approach for the testing and verification of future autonomous driving systems [1]. Their core technologies have become a research focus in the design, manufacturing, and control of complex systems. This paper first systematically reviews the advantages and limitations of current mainstream autonomous driving simulation platforms and further analyzes the necessity of integrating them with digital twin technologies. Subsequently, the digital twin approach is deeply embedded into the autonomous driving research process. A high-fidelity simulation platform is constructed based on real geographic information, and its specific applications in research and experimentation are demonstrated. Experimental results show that the developed digital twin-based autonomous driving simulation platform significantly reduces research costs, simplifies experimental procedures, and improves overall research efficiency.