Resilient distributed model-free adaptive traffic signal control via controller dynamic linearization under DoS...

Area of Science:

• Intelligent Transportation Systems
• Network Security
• Control Theory

Background:

• Traffic signal control faces increasing complexity and network security challenges.
• Growing traffic demands necessitate robust and adaptive control strategies.
• Existing systems are vulnerable to communication disruptions and cyberattacks.

Purpose of the Study:

• To propose a resilient distributed model-free adaptive traffic signal control strategy (CDL-DMFAC).
• To enhance traffic signal control stability and performance under network security threats.
• To mitigate the impact of denial-of-service (DoS) attacks on traffic signal timing.

Main Methods:

• Integration of controller dynamic linearization (CDL) with multi-agent modeling.
• Modeling each signal phase as an independent agent.
• Employing compact form dynamic linearization (CFDL) for controller construction.
• Designing a denial-of-service (DoS) attack compensation mechanism.

Main Results:

• CDL-DMFAC effectively suppresses queue growth and delay accumulation under various attack intensities.
• Performance advantage increases with higher attack severity.
• Achieved 28.3% reduction in average queue length and 36.32% in average waiting time under severe DoS attacks.
• Demonstrated strong resilience and operational stability.

Conclusions:

• The proposed CDL-DMFAC strategy offers significant improvements in traffic flow efficiency and resilience against cyberattacks.
• The method shows potential for deployment in large-scale urban traffic networks.
• Adaptive control integrated with security mechanisms is crucial for future intelligent transportation systems.