Digital Twin Test Method With LTE-V2X for Autonomous Vehicle Safety Test

Abstract

The development of test method is essential for assessing autonomous vehicle safety. How to ensure real vehicle dynamic performance and create more critical scenarios is the challenge for safety test. Therefore, the novel method for autonomous vehicle safety test is necessary. The contribution of this article is twofold. First, the digital twin test method with Internet of Things (IoT) technique is purposed, which seamlessly integrates real test vehicle with virtual environment. We use Vehicle to Everything (V2X) technique to communicate the physical world and virtual environment. The real test vehicle can undergo within the virtual environment, which can create complex critical scenarios conveniently and ensure high-test confidence level. Second, we employ the fault tree analysis method to identify critical scenario element. The identified critical scenario elements can be combined to create various critical scenarios. Furthermore, the case study of autonomous emergency braking (AEB) system in pedestrian crossing collision scenario is implemented with the purposed method. A comparison of the purposed method with traditional methods shows that the purposed method has higher test confidence and more critical scenarios. This study addresses the challenge of safety test with the IoT technique and offers valuable insights for improving autonomous vehicle safety.

Topics & Concepts