Multi-Objective Optimization of Differentiated Urban Ring Road Bus Lines and Fares Based on Travelers’ Interactive Reinforcement Learning

This paper proposes a new multi-objective bi-level programming model for the ring road bus lines and fare design problems. The proposed consists of two layers: traffic management operator travelers. In upper level, we propose fares optimization in which operator’s profit travelers’ utility are set as objective functions. lower evolutionary multi agent bounded rational reinforcement learning with social interaction is introduced. A solution algorithm developed on basis equalization OD matrix. numerical example based real case was conducted to verify models algorithm. computational results indicated that travel choice different degrees rationality significantly changed differentiated fares; furthermore, this can generate reduce maximum section flow, increase profit, generalized cost.