基于选择性深度集成的集装箱吞吐量混合预测模型研究

A hybrid model based on selective deep-ensemble for container throughput forecasting

精准预测集装箱吞吐量是合理规划港口建设、制定港口作业计划和调整港口发展方向的重要基础.针对港口集装箱吞吐量的复杂非线性特征,本文提出了基于选择性深度集成的集装箱吞吐量混合预测模型(HMSD).该模型首先使用经验模态分解方法将原始集装箱吞吐量时间序列分为若干个本征模函数和余波序列.考虑到各本征模函数的高度非线性特征,分别训练长短时记忆网络、门控循环单元和卷积神经网络三种深度学习模型作为基准模型对其进行预测,再运用改进的数据分组处理技术(GMDH)建立选择性深度集成模型,得到各本征模函数的集成预测结果,进一步采用自回归求积移动平均模型预测线性余波序列,将全部本征模函数和余波的预测值进行整合得到总的集装箱吞吐量预测值.为验证提出的模型在集装箱吞吐量预测上的性能,本文选取我国吞吐量具有显著差异的六个港口进行实证,结果表明该模型在六个港口上均具有最好的预测效果,最后还运用HMSD模型给出了2021年1月至2022年12月六个港口集装箱吞吐量的样本外预测值.

The accurate forecasting of container throughput is an important basis for reasonably planning port construction,making port operation plan and adjusting port development direction.Aiming at the complex nonlinear characteristics of port container throughput,a hybrid model based on selective deep-ensemble for container throughput forecasting(HMSD)is presented in this paper.First,this model decomposes the original container throughput time series into several intrinsic mode functions and a residual by empirical mode decomposition.Considering highly nonlinear characteristics of each intrinsic mode function,the proposed model trains three deep neural networks,namely,long short term memory,gated recurrent unit and convolutional neural network,as base learners to predict intrinsic mode functions.Then,this model establishes selective deep-ensemble forecasting model by improved group method of data handling on intrinsic mode functions and obtains their ensemble forecasting results.Furthermore,this model uses an autoregressive integrated moving average model to predict the linear residual.In order to verify the performance of the proposed model in container throughput forecasting,six ports with significant differences in throughput in China are selected for empirical testing,and the results show that the model has the best forecasting effect on all six ports.Finally,the monthly out-of-sample forecasts of container throughput of six ports from January 2021 to December 2022 by HMSD model is given.