Abstract：In order to build satisfactory transportation for the people adapting to the changes of socialist contradictions and adapting to the transformation from "big transportation country" to "strong transportation country", we should strive to improve the efficiency of integrated transportation by means of structural reform on the supply side and comprehensive management on the demand side, so as to realize the transformation of transportation modes from independent development to deep integration. Therefore, starting with the transfer of passenger travel choice, this paper builds a decision-making model of passenger travel behavior based on Bayesian network, in order to further carry out the forecast of passenger travel demand under the background of the integration of multiple modes of transportation. Firstly, according to the theory of consumer surplus, the basic workflow of travel mode selection is constructed under the criterion of maximizing consumer surplus. Secondly, according to the relationship among the influencing factors, the Bayesian network node is set as the influencing factor of travel mode selection. Thirdly, assigning values to each node and transforming them into joint trees based on conditional probability. Finally, the Bayesian polynomial Logit model for passenger travel mode selection is established, and the actual situation is discussed and analyzed to obtain better model estimation results. This study would help to better understand the influencing factors of passenger travel mode choice decision-making behavior, and achieve accurate prediction of traffic demand, so as to improve the efficiency and effect of integrated transportation resources allocation in the context of the new stage. Keywords：Decision-making Behavior of Passenger Travel Mode；Bayesian Network；Integrated Transportation；

CICTP