摘要：AbstractStrategic traffic management is of grave importance to combat traffic congestion at macroscopic level. However, such a field is still relatively less-explored, particularly in engineering practice. This paper defines traffic operational mode with its switch process and proposes an agent-based recommendation framework accordingly. Such a framework employs a deep neural network architecture and is capable of recommending signal cycle length at intersection-level to the traffic operators upon operational mode switch.
关键词：KeywordsAgent-based systemstrategic traffic managementdeep learningrecommendation system
