基于IOWA算子的港口吞吐量组合预测方法

Combined prediction method for port throughput based on induced ordered weighted averaging operator

为了合理地预测港口吞吐量,给出了一种基于诱导有序加权平均(IOWA)算子的港口吞吐量组合预测方法。建立了基于IOWA算子的港口吞吐量组合预测模型;选择具有较强非线性映射能力的BP神经网络和能够处理少量不完全信息的GM(1,1)作为单项预测模型;最后以连云港港口吞吐量为例,对预测效果的准确性进行了评价并科学合理地预测了连云港2013—2015年的港口吞吐量。结果表明,基于IOWA算子的组合预测方法预测效果的准确性较高。

In order to rationally predict port throughput, a combined prediction method based on the induced ordered weighted averaging( IOWA) operator is proposed.Firstly, a combined prediction model for the port throughput based on IOWA operator is established.Secondly, the BP network which has strong nonlinear mapping ability and GM(1, 1) which can handle small amount of incomplete information are regarded as the prediction models for single item. Finally, through a case study of Lianyungang Port, the accuracy of predicted results is evaluated, and the predicted throughput of Lianyungang Port from 2013 to 2015 is given. A conclusion is drawn that the accuracy of the predicted results by the combined prediction model based on IOWA operator is satisfactory.