大丰港货物吞吐量的时间序列法预测研究

Prediction Based on Time Series Method of Cargo Throughput of Dafeng Port

根据大丰港的历史数据,采用时间序列的回归分析方法对大丰港货物吞吐量进行预测研究。在回归分析法中,通过比较,选择指数模型、高次多项式模型以及支持向量机方法分别对大丰港的货物吞吐量进行模拟预测。对指数模型、多项式模型、支持向量机三种方法的预测结果进行比较,并结合大丰港的实际情况,最终选择使用支持向量机法给出大丰港2012—2016年的货物吞吐量预测值。

Based on time series method,the regression analysis method is used to predict the cargo throughput of Dangfeng port.In the regression analysis method, the exponential function model,the polynomial model and the support vector machine are used respectively through comparison to make a prediction. By contracting the results of three methods, the result of the support vector machine is more reasonable and reliable. Finally, the prediction of cargo throughput of Dafeng port from year 2012 to 2016 is given by the support vector machine.