摘要
为了提高船舶流量预测精度,综合考虑交通需求、季节、气候等因素,通过分析船舶流量历史数据,在线性增长模型基础上,构建了考虑周期性波动因素的船舶流量预测改进模型,并运用贝叶斯估计和预测方法求解模型,提出了一种新的船舶流量预测方法.实例验证表明,与传统线性增长模型预测结果比较,新模型更符合船舶流量实际情况,月流量预测结果的平均绝对误差下降了3.56%,标准差下降了3.79%.因此,将该预测方法用于船舶流量预测是有效的.
In order to improve prediction accuracy of ship flow,an improved linear growth model is developed to predict ship flow by taking into all periodic fluctuation factors,such as actual ship demand,seasonal changes,climate impact,and so on. Then,the Bayesian estimation and prediction are used to solve the new model,and ship flow is predicted using the historical data of ship flow. A case is analyzed to compare the prediction effect of the model,and the results show that,compared with the linear growth model,the prediction results of the improved model are more in line with the actual situation of ship flow. Besides,the mean absolute error of monthly ship flow decreases by 3. 56%,and the standard deviation decreases by 3. 79%. Therefore,the method is effective to predict ship flow.
出处
《江苏科技大学学报（自然科学版）》
CAS
北大核心
2017年第4期531-536,共6页
Journal of Jiangsu University of Science and Technology:Natural Science Edition
基金
江苏省高校哲学社会科学研究课题(2015JSB326)
关键词
船舶流量
周期性波动
线性增长模型
预测
ship flow
periodic fluctuation
linear growth model
prediction
