Guidance compliance behaviors of drivers under different information release modes on VMS

• SOAR is first employed and designed to study guidance compliance behavior.
• Working memory and long-term memory of drivers are represented distinctively.
• Chunking etc. innovate drivers’ learning mechanism.
• Accurate simulations enhance the effectiveness of the conclusions.

Driver’s guidance compliance behavior, which is a crucial link in Intelligent Traffic Guidance Systems (ITGS), has a direct impact on guidance effect. Based on State, Operator, and Result (SOAR), which was a cognitive architecture, the traffic compliance agent was designed and elaborated in detail. In the paper, working memory (WM), long-term memory (LTM), operator selection, impasse solving, expected travel time determination, and chunking rule generation were described with the consideration of practical situations. Finally, two simulations with different information release modes were carried out under the same given condition, one was based on real-time flow statistics (M1), and the other was based on flow forecasting (M2). Through analyzing the changes of drivers’ compliance rates and vehicles’ lane-changing times under Variable Message Sign (VMS), M2 was proven to be more effective in alleviating traffic jam in the morning and evening peak periods and bring a higher compliance rate. This study laid a foundation for selecting the release modes of guidance information in both theory and practice.