Connected and automated vehicles (CAVs) are recognized to have the great potential to alleviate traffic congestion and improve traffic safety, and thus be regarded as the direction of future transportation. The assurance of autonomous driving system safety in critical situations is a challenging task, and the testing methods for traditional vehicles can no longer meet the requirements of CAVs. According to assessments, CAVs need at least 240 million kilometers of tests without any accidents to prove th

Connected and automated vehicles;Generative adversarial network;Safety performance evaluation