Research on Transportation Mode Split Model Based on BP Neural Network Between Highway And Railway in Transportation Corridor

摘要

In a transportation corridor involving highway and railway, research on transportation mode split is very important both theoretically and practically for regulating the passenger transportation market, improving the transportation level of service, and accommodating the increasing travel demand.Logit model which is a kind of stochastic discrete model is widely used in transportation mode split By analyzing the shortcomings of Logit model utility function, the paper discusses how to solve the problem of transportation modes split by using BP Neural Network.The paper puts forward Logit model based on BP Neural Network which improves the algorithm of the utility value based on neural network nonlinear approximation ability.Then, with an example, the paper expatiates how to establish and train transportation mode split model based BP nerve network and Matlab Toolbox.Result tests the accuracy and feasibility of the model based on BP Neural Network and indicates that model has a better utility value.