摘要
由于港口吞吐量的影响因素相当多,这些影响因素中大部分又是不可量化指标,造成数据收集和确定的困难。在介绍BP算法的基础下,使用基于时间序列的BP神经网络模型对防城港货物吞吐量进行预测,该模型不仅解决影响因素多、数据难收集的问题,而且是货物吞吐量预测方法中精度很高的一种有效方法。
There are many factors that influence the port throughput, and the majority of these factors is not quantifi- able indicators, resulting in data collection and identification difficult. This paper, based on BP network, uses time-series- based BP network model to predict Fangchenggang cargo throughput. The model not only solves the problem of the dif- ficulty of data collection and identification, but also is one effective method of cargo throughput.
出处
《物流科技》
2010年第5期33-35,共3页
Logistics Sci-Tech
