Abstract: Taking London as a case study, the causal effect method of virtual synthetic control group was used to estimate the impact of congestion pricing on traffic accident casualties in central London. By comparing the actual number of casualties in London and the number of casualties estimated by the virtual synthetic control group, the total number of casualties in traffic congestion toll area decreased by 4.29%, in which the number of minor injuries and serious injuries decreased by 5.05% and 12.12% respectively. The proposed model was validated by placebo test. The results show that the evaluation effect of congestion charge in London reaches 95% confidence level. The results show that the virtual synthetic control method can be used as an effective research method when there is no suitable comparative control group in traffic safety analysis.

Key words: traffic and transportation engineering, London congestion charge, virtual synthetic control group, causal effect, traffic policy, placebo-controlled study

摘要： 以伦敦为研究案例，采用虚拟对比控制组的因果效应方法，估计伦敦中心区实施拥堵收费对交通事故伤亡人数的影响。通过对比伦敦实际发生的伤亡人数和虚拟对比控制组估计的伤亡人数，分析发现交通拥堵收费区的交通事故总伤亡人数减少了4.29%，其中轻伤和重伤数量各自减少了5.05%和12.12%。采用安慰剂试验对模型进行验证，结果显示伦敦交通拥堵费的评估效应达到95%的置信水平。研究结果表明：在交通安全分析研究中，当没有合适的对比控制组的情况下，虚拟对比控制方法可作为一个有效的研究方法。

关键词: 交通运输工程, 伦敦交通拥堵费, 虚拟对比控制组, 因果效应, 交通政策, 安慰剂试验