A BP-based path selection model for dynamic transportation network

Abstract

This paper concentrates on the issues commonly existed in the dynamic transportation network, for instance, there are large number of stochastic situation, time-oriented cases and it is difficult to work out the optimal path. In order to deal with these issues, a BP (Back Propagation)-based path selection model is introduced for the dynamic transportation network. The model utilizes neuron model together with two training and learning methodologies like LSA and EBP to improve the velocity of convergence and reduce the running time. Experiments are carried out to compare the traditional algorithm with this model proposed in this paper. The simulation results imply that the proposed model is better than the traditional algorithm in terms of training performance.