| Rail freight demand forecasting as the basis of railway freight transportation system planning,is for freight equipment and efficient allocation of resources and the construction of the important basis of efficient freight system,in order to make railway planning work more targeted and guiding significance to the research for railway freight demand forecasting theory and the method is particularly important.Based on the research status of railway freight demand forecasting at home and abroad and relevant theoretical knowledge,this paper combines the characteristics of railway freight demand and takes China railway as the background to carry out the research on railway bulk cargo transport demand forecasting.I mainly did the following work:(1)Analyze the sources and influencing factors of bulk cargo demand,and build a forecasting index system for bulk cargo demand.(2)Aiming at the problem that the standard particle swarm optimization algorithm is difficult to balance global search and local mining and easy to get into local optimization,the standard particle swarm optimization algorithm is improved.(3)According to the characteristics of bulk cargo transportation system combined with the grey model weakening data randomness,and the characteristics of the neural network nonlinear fitting capability is strong,the use of the improved particle swarm optimization(pso)algorithm on gray neural network structure and individual parameters optimization,set up the improved particle swarm optimization grey neural network prediction model is used to solve the bulk cargo transport demand forecasting in a given area.(4)According to the research on the characteristics of the existing transportation mode in China,the utility function is established on the basis of the factors influencing shipper’s choice of cargo transportation mode,and the MNL model is used to solve the problem of calculating the route’s share of cargo transportation,and the improved particle swarm optimization algorithm is used to estimate the parameters in the model.(5)Make an empirical study on the coal transport demand forecast of Watang-Rizhao Railway under the background of "west coal transport to east coal transport".Empirical study shows that by the improved particle swarm optimization algorithm in the initial weights and threshold of gray neural network model and network structure optimization,promotes the stability and accuracy of the prediction model,the average relative error is only 1.86%,and the improved particle swarm optimization(pso)algorithm and estimate its parameters,to share the volume of the calculation model,finally calculate the average error of 4.9%,the share rate so that based on improved particle swarm optimization(pso)algorithm to estimate parameters of rail bulk cargo transportation share the quantity calculation model is feasible. |
