Towards robust car-following based on deep reinforcement learning. Transportation Research Part C: Emerging Technologies, (159):104486, 2024. [PUMA: topic_engineering Car-following; Generalization; Reinforcement; String Validation; capabilities; learning; model; stability]