关键词: 铁路货运量预测, 粒子群优化算法, 灰色神经网络, 灰色关联分析, BP神经网络, Elman神经网络

Abstract: Concerning the shortcomings of the methods which forecast railway freight volume, the paper proposed Grey Neural Network (GNN) based on the Improved Particle Swarm Optimization algorithm (IPSO-GNN). To make up for the shortfall of the conventional GNN and guarantee the prediction accuracy, it optimized the GNN whitening parameters through the IPSO. And it computed the railway freight volume and the correlation degree of influential factors. It built a railway freight volume model based on IPSO-GNN with six relating factors. The simulation results show that the prediction method is effective and feasible. The prediction precision of the given model in the railway freight volume forecast is better than those of the conventional GNN prediction method and other prediction methods.

Key words: railway freight volume prediction, Particle Swarm Optimization (PSO) algorithm, Grey Neural Network (GNN), grey relation analysis, BP neural network, Elman neural network