Modelling herd behaviour in traffic jams using Markov chains-based reinforcement learning learning

Key Points

• This research aims to model herd behaviour during traffic jams using Markov chains and reinforcement learning techniques.
• Applied Markov chains to represent traffic flow dynamics.
• Utilized reinforcement learning to optimize traffic behaviour responses.
• Simulated various traffic scenarios to study herd effects.
• Demonstrated that herd behaviour significantly impacts congestion levels.
• Reinforcement learning improved prediction accuracy of traffic patterns.
• Identified key factors influencing driver interactions during traffic jams.

Abstract