Simulation Modeling and Application of Travel Mode Choice Based on Bayesian Network

摘要

In this paper, we study the travel mode choice of residents to determine the set of factors which can influencetravel mode choice of residents and analyze the influence factor characteristics. Using Bayesian theory, we analyze thetravel decision-making data of the residents, discrete them, and use them in Bayesian network structure learning andparameter estimation by K2 algorithm. We establish a Bayesian network simulation model to analyze the dependenceprobability relationship between the parent nodes and child nodes. Validation test was carried out for the buildingsimulation model of Bayesian network. Data analysis results showed that the Bayesian network has a high accuracyprediction for actual travel mode choice of residents. This paper studies the Bayesian structure and parameters learningmethod for the actual travel behavior, and this method which provides a new method for studying the travel mode choiceof residents can reveal the relationship between the various attributes associated with travel mode choice through a newperspective.