愤怒驾驶在全球非常普遍，通常会导致危险驾驶行为，甚至严重的交通事故。 因此，探讨愤怒情绪背后的影响特征，并据此提出相应的干预措施显得尤为重要。 对于这一驾驶领域的传统研究问题，本文设计了新的研究方案，以特征探索、干 预对比以及应用设计为主线，开展了四个实验研究，并利用脑电信号测量等手段， 探索了愤怒驾驶员的风险行为和认知特征，提出了相应的干预措施和应用设计指 导。结合本文研究成果，未来可利用人工智能等作为技术支撑，开展情绪干预。 在特征探索阶段，本文首先采用驾驶仿真模拟的方法，对愤怒状态下的风险 行为进行阶段划分。研究结果表明，愤怒驾驶员面对风险时采取险情处置的距离 和时间均较长，其风险应对能力下降，导致超速、碰撞等危险驾驶行为的发生。然 后，本文利用脑电研究方法，结合黄灯信号灯 “走” 或 “停” 的驾驶场景，设计了 经典的风险决策实验。研究发现驾驶员在愤怒状态比平静状态下具有更高的 P300 波幅，并且在高风险决策时也呈现出比低风险更高的 P300 波幅。结果表明，在认 知决策过程中，他们对高风险更加敏感，对报酬激励有更强的动机，期望通过高 风险决策来获得高报酬，导致了后续高风险行为的增加。研究表明愤怒情绪会通 过影响脑认知资源投入，调整报酬激励评估和风险评估，进而影响风险决策过程。 在干预对比阶段，本文结合上述研究内容，设计了基于正负反馈内容的愤怒 干预措施，并在驾驶仿真场景中对比其有效性。研究结果表明，与负向干预相比， 正向干预能够显著缓解驾驶员的愤怒情绪、降低脑力负荷、增强驾驶控制并降低 危险驾驶行为，同时具有更高的主观评价。与正向激励相关的干预内容，可以为 驾驶绩效的提升提供帮助，并且可以对车载愤怒干预系统的设计提供启发。 在应用设计阶段，本文进一步对多种干预方式进行评估，寻找有效的干预措 施，为车载愤怒干预系统提供设计指导。本文采用焦点小组访谈和问卷调研的方 法，对车载愤怒干预系统的潜在需求进行分析，发现驾驶员最喜欢正向反馈的干 预方式。同时，一些与驾驶相关的人格特质与干预类型的偏好或对干预系统的态 度存在一定关联，为车载愤怒干预系统的设计提供了新的视角。 本文的研究结果是对愤怒驾驶相关研究领域的深入和扩展，对愤怒驾驶的行 为和认知特征进行了研究探索，并为愤怒干预措施的设计提供了新的视角，研究 结果可以辅助驾驶员进行安全驾驶。

Angry driving is very common around the world and often leads to dangerous driving behaviors and even serious traffic accidents. Therefore, it is of great importance to explore the characteristics of the influence underlying angry driving and then propose corresponding interventions. Based on this classical research question in the driving field, this dissertation designed a new research program from characteristic exploration, intervention comparison to application design. This dissertation conducted four experimental studies that used EEG measurement to explore the characteristics of angry drivers’ risky behavior and cognition, and proposed corresponding intervention methods as well as de- sign implications. Meanwhile, combined with the results obtained, artificial intelligence can be used as technical support to assist emotional intervention design in the future.In the stage of characteristic exploration, this dissertation employed a driving simulation method and divided the risky behaviors into phases under anger. The results showed that angry drivers took more time and distance when facing risk. Their ability of reaction to risk was decreased, leading to dangerous driving behaviors such as speeding and collisions. Then, this dissertation used the neuroscience method to design a classical risky decision-making experiment according to the driving scenario of “Go” or “No-Go” before a yellow traffic light. Results showed that angry drivers had higher P300 amplitudes than they were in the baseline state and also higher P300 amplitudes in high-risk decisions than low-risk decisions. The results indicated that they were more sensitive to high risks and had higher motivation for rewards, and expected high rewards through high-risk choices, which led to an increase in subsequent high-risk behaviors. The results demonstrated that anger could influence the decision-making process by affecting cognitive resource involvement, rewards evaluation, and risk assessment.In the stage of intervention comparison, based on the above studies, this dissertation designed the interventions according to positive and negative feedback content, and compared the effectiveness of different intervention methods in the simulated driving scenarios. The results showed that positive interventions can significantly mitigate drivers’ anger, reduce mental workload, enhance driving control and reduce dangerous driving behavior, as well as obtain higher subjective evaluations than negative interventions. It in- dicated that the intervention content related to positive feedback can help improve drivingperformance and provide implications for the design of the in-vehicle anger intervention systems (AISs).In the stage of application design, this dissertation evaluated various anger interventions to find effective interventions, and provide design implications for in-vehicle AISs. This paper conducted focus group interviews and questionnaire surveys to analyze the potential requirements of AISs. Results showed that drivers preferred positive feedback as an intervention method. Meanwhile, some personality traits related to driving anger had associations with the preference of interventions or the attitudes of the AISs, which provided a new perspective for the design of the in-vehicle AISs.The results of this dissertation helped deepen and expand the angry driving research domain, by exploring the influence characteristics of angry driving and providing a new perspective for the design of AISs. The results of this research can help drivers drive safely.