摘要：用于 ADAS 评测的轨迹数据集严重偏向常规驾驶场景；真实的车车冲突事件十分罕见，而事件越罕见，ADAS 系统未能妥善处理时的代价就越高。现有生成方法通过场景级属性——空间目标、智能体结构或自然语言对抗目标——进行条件控制，以应对这种不平衡；但没有方法能够将目标碰撞时间（Time-to-Collision，TTC）值作为输入，并能够以可测量误差为标准确保生成该目标值。本文提出 SevDiff：一种以严重度为条件的去噪扩散概率模型（DDPM）。该模型将请求的最小 TTC 值作为标量条件信号输入，生成成对车辆交互轨迹；通过命中率指标评估其实际产生的冲突严重度是否符合请求。SevDiff 在从 UTE SQM-W-1 高速公路交织路段数据集提取的 468 个交互窗口上训练，该数据集经平滑处理后包含 1,041 辆车辆和 822,691 条观测记录。对于 0.5–1.5 秒的 TTC 目标，在 ±0.5 秒范围内，SevDiff 的命中率达到 100%；对于 2.0–2.5 秒目标，命中率为 97%–99%；当 TTC = 5.0 秒时，性能平稳下降至 39%。生成轨迹的运动学特征具有物理合理性：在 12 项特征中，超出合理范围的最高比例为 4.7%；超过 96.5% 的样本不存在负速度或负间距值。命中率的下降模式可从物理角度解释为：条件信号相对于训练先验的强度发生变化。因此，这一结果是对生成器精度的刻画，而非简单的通过/失败判定。

Abstract:Trajectory datasets used in ADAS evaluation are heavily biased toward routine driving; genuine vehicle-to-vehicle conflict events are rare, and the rarer the event, the higher the cost when an ADAS system fails to handle it. Existing generative approaches address this imbalance by conditioning on scene-level properties - spatial goals, agent structure, or natural-language adversarial objectives - but none can accept a target Time-to-Collision (TTC) value as input and be held to producing it within a measurable error. This paper introduces SevDiff, a severity-conditioned denoising diffusion probabilistic model (DDPM) that accepts a requested minimum TTC value as a scalar conditioning signal and generates paired vehicle interaction trajectories whose realized conflict severity matches the request, evaluated through a hit-rate metric. Trained on 468 interaction windows extracted from the UTE SQM-W-1 expressway weaving-section dataset (1,041 vehicles, 822,691 observations after smoothing), SevDiff achieves 100% hit-rate within +/-0.5 s for TTC targets of 0.5-1.5 s and 97-99% at 2.0-2.5 s, with graceful degradation to 39% at TTC = 5.0 s. Generated kinematic features are physically plausible, with a maximum out-of-range rate of 4.7% across 12 features and no negative speed or gap values in more than 96.5% of samples. The hit-rate degradation pattern is physically interpretable as the strength of the conditioning signal relative to the training prior, making it a precision characterization of the generator rather than a pass/fail result.