基于自适应过滤法的香港货运海船交通量预测
Hong Kong Ocean Cargo Ship Traffic Flow Prediction Based on the Adaptive Filtering Method
谢佳仪 1陈丽宁 1曾烁楠1
作者信息
- 1. 广州航海学院 海运学院,广东 广州 510725
- 折叠
摘要
为准确预测香港货运海船月交通量,建立了基于自适应过滤法的香港货运海船交通量预测模型.建模所用的统计数据来源于香港特别行政区海事处网站.首先直接使用自适应过滤法建模,进而使用数据标准化的自适应过滤法建模.比较这两个模型发现,后者的均方差更小、平均相对误差更小,两者预测时间序列与样本时间序列的灰色绝对关联度接近,这意味着后者的离散性、精度均优于前者,预测时间序列与样本时间序列折线的几何相似性与前者接近.因此,用数据标准化的自适应过滤法模型预测了未来8 个月香港货运海船交通量.
Abstract
To accurately predict Hongkong ocean cargo ship traffic flow,the adaptive filtering method is used to establish the prediction model for Hongkong ocean cargo ship traffic flow.The statistical data of model establishment is gained from the web of Marine Department,the Government of Hong Kong SAR.Firstly,the adaptive method is directly used to establish the prediction model,and then the adaptive method with data normalization is applied.After comparison of the two models,it is found that the later has smaller mean square error,lower mean relative error,and the two models have close grey absolute relational grade between the prediction time series and sample time series.It means that the later model performs better in the discreteness and accuracy than the former,and the two models have similar geometrical similarity between the curve of the prediction time series and that of the sample time series.So,the adaptive method with data normalization is applied to predict Hong Kong ocean cargo ship traffic flow in future 8 months.
关键词
香港/货运海船/交通量/自适应过滤法Key words
Hong Kong/ocean cargo ship/traffic flow/adaptive filtering method引用本文复制引用
基金项目
广东省高等教育教学改革项目(C2206001114)
出版年
2024
NETL
NSTL
万方数据