基于因素分析和SVM模型的江海联运运量预测方法

鉴于江海联运运量受众多因素的影响, 为了解决江海联运运量预测问题, 先对江海联运运量影响因素进行分析, 再用灰色关联度分析筛选出其中的典型因素. 在此基础上应用粒子群算法优化的支持向量机建立江海联运运量预测模型, 应用于宁波港域江海联运量的预测. 结果表明, 该模型与传统时间序列预测方法相比具有较高的拟合度和预测精度, 为解决江海联运运量等多因素非线性系统预测提供了一条新的途径.

Method for prediction of river-sea combined transport volume based on factor analysis and SVM model

The volume of river-sea combined transport is affected by many factors. In order to solve the problems found in forecasting this volume, its influencing factors are analyzed first, and then typical factors are selected by grey relation analysis. Based on the result, a support vector machine (SVM), optimized by particle swarm optimization (PSO) algorithm, is applied to establish a predictive model of the river-sea combined transport volume. The model is then applied to the prediction of the river-sea combined transport volume in Ningbo Port. The results show that the model is superior to the traditional time series forecasting method, and has a high degree of fitness and forecasting accuracy. It provides a new way to solve the multi-factor nonlinear system prediction problems, such as the one in river-sea combined transport volume prediction.