关键词: 交通运输工程；港口吞吐量；组合预测；分数阶累加；支持向量回归；诱导有序加权平均算子

Abstract: In order to improve the prediction accuracy of port throughput, a combined prediction model based on fractional order cumulative forecast GM (1,1) prediction model FGM (1,1) and support vector regression (SVR) was established to predict the port throughput. Firstly, FGM (1,1) model and SVR model were applied to predict the port throughput respectively. Then, aiming at the problem that the traditional combined model weighting could not take into account the predictive power of each individual model at each time point, a weighting method based on of induced ordered weighted averaging (IOWA) operator was proposed for combination prediction. Finally, taking the cargo throughput of Chongqing Port from 2005 to 2020 as the data sample for example verification, the FGM (1,1) model, SVR model and weighted combined model were used for port throughput prediction, respectively. The prediction accuracy of the three models were compared and these three models were used to respectively predict port throughput from 2021 to 2024. The research results show that the root mean square error, mean absolute error and mean absolute percentage error of the combined model based on IOWA are significantly smaller than those of each single prediction model．The proposed combined model could provide a new method for port throughput prediction.

Key words: traffic and transportation engineering; port throughput; combination prediction; fractional order cumulative; support vector regression; induced ordered weighted averaging operator