Evaluation of net resilience effects of signal control failure events in mixed traffic flow

ZHANG Haosen, WANG Jingsheng, RAO Jianghua, ZOU Xinxu

Journal of Tsinghua University(Science and Technology) ›› 2026, Vol. 66 ›› Issue (11) : 2285-2298.

PDF(6356 KB)
PDF(6356 KB)
Journal of Tsinghua University(Science and Technology) ›› 2026, Vol. 66 ›› Issue (11) : 2285-2298. DOI: 10.16511/j.cnki.qhdxxb.2026.26.067
OPERATIONAL SAFETY ASSURANCE FOR AUTONOMOUS ROAD TRAFFIC SYSTEMS

Evaluation of net resilience effects of signal control failure events in mixed traffic flow

  • {{article.zuoZhe_EN}}
Author information +
History +

Abstract

[Objective] With the development of connected and automated vehicle (CAV) technology, urban road networks increasingly carry mixed traffic flow composed of CAVs and human-driven vehicles. Although higher CAV penetration improves efficiency and stability under normal conditions, these benefits cannot be directly interpreted as resilience gains under disruptive events. Focusing on signalized networks affected by signal control failure events, this study separates the normal-operation baseline effect from the additional net resilience gain, identifies penetration thresholds, and clarifies whether CAVs mainly resist the initial impact or promote post-failure recovery. [Methods] A SUMO-based microscopic simulation experiment was developed for the grid-like road network of the Suzhou Xiangcheng intelligent connected vehicle demonstration area, with road topology, signal timing, boundary inflows, turning flows, and road attributes calibrated using multisource data. Eleven CAV penetration levels were combined with four failure levels to form the scenario matrix, in which L0 denotes the no-failure baseline and L1, L2, and L3 represent single-intersection, corridor-level, and area-level failures, respectively. The failure was injected at 3 600 s and lasted 1 800 s, and the process was divided into warm-up, baseline, failure, and recovery periods to capture delayed restoration of the network state. Nested vehicle replacement and common random numbers were adopted to improve comparability. The average vehicle time loss served as the core outcome, with the average speed, stopped vehicles, queue exposure, recovery residuals, and completion rate used for auxiliary verification. Based on the potential-outcomes notation, static difference-in-differences, event-study dynamic difference-in-differences, and stage-based difference-in-difference-in-differences were integrated to remove the baseline effect, trace dynamic net effects, and decompose stage-specific mechanisms, with block bootstrap resampling and false discovery rate correction used for inference. [Results] CAV penetration produced clear normal-operation benefits even without signal failure: in L0, the reduction in average vehicle time loss increased from 0.03 to 0.79 s/veh with rising penetration. After this baseline effect was removed, the additional net resilience gain showed evident nonlinearity and failure-level heterogeneity, stabilized at approximately 60% under L1, 40% penetration under L2, and 50% under L3. At full penetration, the overall gains under L1, L2, and L3 reached approximately 1.00, 1.15, and 1.29 s/veh, respectively. Marginal net gains were concentrated mainly in the 50%-70% interval, with the L3 peak reaching approximately 0.14 s/veh. Dynamic and stage-decomposed results further indicated that net gains arose mainly in the late failure stage and the recovery stage. At full penetration, the recovery-stage net effects under L1, L2, and L3 were 0.302, 0.538, and 0.718 s/veh, respectively, markedly larger than the corresponding failure-stage effects of 0.077, 0.120, and 0.128 s/veh. Auxiliary indicators confirmed this recovery-dominant mechanism in terms of average speed, stopped vehicles, queue exposure, recovery residuals, and completion rate. Robustness checks showed that common-random-number pairing reduced seed-level variance by more than 50%, that alternative statistical windows preserved the main dynamic patterns and that perturbations of key time headways changed neither the thresholds nor the positive net gains. [Conclusions] CAV effects under signal control failure are evaluated by distinguishing normal-operation efficiency benefits from additional net resilience gains. The proposed framework provides a transparent approach for quantifying net gains and decomposing their temporal and stage-specific mechanisms. CAVs enhance the resilience of signalized urban road networks mainly by strengthening recovery rather than by fully offsetting the initial failure impact. These findings support differentiated CAV deployment and emergency management in mixed traffic systems, although further validation is needed for other network structures, demand levels, failure durations, and failure modes.

Key words

mixed traffic flow / paired-difference estimation / difference-in-differences method / macroscopic fundamental diagram / net resilience gain

Cite this article

Download Citations
ZHANG Haosen, WANG Jingsheng, RAO Jianghua, ZOU Xinxu. Evaluation of net resilience effects of signal control failure events in mixed traffic flow[J]. Journal of Tsinghua University(Science and Technology). 2026, 66(11): 2285-2298 https://doi.org/10.16511/j.cnki.qhdxxb.2026.26.067

References

[1] 新华社. 中华人民共和国国民经济和社会发展第十五个五年规划纲要[N]. 人民日报, 2026-03-14(1). Xinhua News Agency. Outline of the Fifteenth Five-Year Plan for National Economic and Social Development of the People’s Republic of China[N]. People’s Daily, 2026-03-14(1). (in Chinese)
[2] 李帅, 王嘉伟, 许庆, 等. 网联环境下混合交通流稳定性分析[J]. 交通信息与安全, 2025, 43(6): 76-85. LI S, WANG J W, XU Q, et al. A stability analysis of mixed traffic flow in connected environment[J]. Journal of Transport Information and Safety, 2025, 43(6): 76-85. (in Chinese)
[3] 张鲁宁, 王景升. 混合交通流异常事件下自动驾驶车辆轨迹优化[J/OL]. 重庆大学学报. (2025-06-25) [2026-03-30]. https:// link.cnki.net/urlid/50.1044.N.20250625.1513.002. ZHANG L N, WANG J S. Optimization of autonomous vehicles trajectories under abnormal events in mixed traffic flow[J/OL]. Journal of Chongqing University. (2025-06-25) [2026-03-30]. https://link.cnki.net/urlid/50.1044.N.20250625.1513.002. (in Chinese)
[4] 宗芳, 李宇暄, 曾梦, 等. 考虑相变滞后效应的交通状态估计与速度预测[J]. 中国公路学报, 2025, 38(8): 171-186. ZONG F, LI Y X, ZENG M, et al. Traffic state estimation and traffic flow speed prediction considering hysteresis effect[J]. China Journal of Highway and Transport, 2025, 38(8): 171-186. (in Chinese)
[5] LI J J, YU C H, SHEN Z L, et al. A survey on urban traffic control under mixed traffic environment with connected automated vehicles[J]. Transportation Research Part C: Emerging Technologies, 2023, 154: 104258.
[6] 杜若熙. 混合交通流条件下的城市道路交通网络系统韧性研究[D]. 北京: 北京交通大学, 2023. DU R X. Research on traffic resilience of urban traffic network system under mixed traffic flow condition[D]. Beijing: Beijing Jiaotong University, 2023. (in Chinese)
[7] 刘庆, 龚华凤, 何易非, 等. 城市道路交通网络韧性研究进展[J]. 城市交通, 2025, 23(4): 9-24. LIU Q, GONG H F, HE Y F, et al. A review of the resilience of urban road transportation network[J]. Urban Transport of China, 2025, 23(4): 9-24. (in Chinese)
[8] 卫亚欣, 李堃, 慕晨, 等. 城市韧性交通系统综述: 评估方法与优化提升[J]. 交通运输工程学报, 2026, 26(4): 230-258. WEI Y X, LI K, MU C, et al. Review of urban resilient transportation systems: Assessment methods and optimization improvements[J]. Journal of Traffic and Transportation Engineering, 2026, 26(4): 230-258. (in Chinese)
[9] LIU Z Z, CHEN H, LIU E Z, et al. Exploring the resilience assessment framework of urban road network for sustainable cities[J]. Physica A: Statistical Mechanics and its Applications, 2022, 586: 126465.
[10] PEREZ Y, PEREIRA F H. Simulation of traffic light disruptions in street networks[J]. Physica A: Statistical Mechanics and its Applications, 2021, 582: 126225.
[11] CHEN G Z, VAN WOENSEL T, XU J H, et al. Assessing movement-specific resilience of a signalized road network under lane-level cascading failure[J]. Physica A: Statistical Mechanics and its Applications, 2024, 654: 130154.
[12] LU Q L, SUN W Z, DAI J N, et al. Traffic resilience quantification based on macroscopic fundamental diagrams and analysis using topological attributes[J]. Reliability Engineering & System Safety, 2024, 247: 110095.
[13] 张寅, 帅斌, 范成敬, 等. 基于行车风险脉冲场的混合交通流模型[J]. 公路交通科技, 2025, 42(12): 1-10. ZHANG Y, SHUAI B, FAN C J, et al. Mixed traffic flow model based on driving risk pulse field[J]. Journal of Highway and Transportation Research and Development, 2025, 42(12): 1-10. (in Chinese)
[14] 秦严严, 王昊, 王炜, 等. 混有CACC车辆和ACC车辆的异质交通流基本图模型[J]. 中国公路学报, 2017, 30(10): 127-136. QIN Y Y, WANG H, WANG W, et al. Fundamental diagram model of heterogeneous traffic flow mixed with cooperative adaptive cruise control vehicles and adaptive cruise control vehicles[J]. China Journal of Highway and Transport, 2017, 30(10): 127-136. (in Chinese)
[15] 秦严严, 王昊, 冉斌. CACC车辆跟驰建模及混合交通流分析[J]. 交通运输系统工程与信息, 2018, 18(2): 60-65. QIN Y Y, WANG H, RAN B. Car-following modeling for CACC vehicles and mixed traffic flow analysis[J]. Journal of Transportation Systems Engineering and Information Technology, 2018, 18(2): 60-65. (in Chinese)
[16] 钱勇生, 杨紫龙, 曾俊伟, 等. 考虑混入自动驾驶车辆的机非混合交通流元胞自动机仿真[J]. 计算机工程与应用, 2026, 62(5): 383-392. QIAN Y S, YANG Z L, ZENG J W, et al. Cellular automaton simulation of mixed traffic flow with autonomous vehicles[J]. Computer Engineering and Applications, 2026, 62(5): 383-392. (in Chinese)
[17] GUO Y M, LI M, LI K Q, et al. Unraveling the determinants of traffic incident duration: A causal investigation using the framework of causal forests with debiased machine learning[J]. Accident Analysis & Prevention, 2024, 208: 107806.
[18] GRAHAM D J. Causal inference for transport research[J]. Transportation Research Part A: Policy and Practice, 2025, 192: 104324.
[19] CAO D N, WU J J, DONG X L, et al. Quantification of the impact of traffic incidents on speed reduction: A causal inference based approach[J]. Accident Analysis & Prevention, 2021, 157: 106163.
[20] DONG P P, ZHANG X N, ZHANG X G. A data-driven approach for spatio-temporal causal analysis in large-scale urban traffic networks[J]. Transportation Research Part E: Logistics and Transportation Review, 2025, 202: 104244.
[21] MILANÉS V, SHLADOVER S E. Handling cut-in vehicles in strings of cooperative adaptive cruise control vehicles[J]. Journal of Intelligent Transportation Systems, 2016, 20(2): 178-191.
[22] YU W J, HUA X D, ZHAO D, et al. Safety impact of cooperative adaptive cruise control vehicles’ degradation under spatial continuous communication interruption[J]. IET Intelligent Transport Systems, 2022, 16(3): 309-331.
PDF(6356 KB)

Accesses

Citation

Detail

Sections
Recommended

/

〈 〉