摘要
货物运输量预测对交通运输基础设施建设和区域经济发展具有重要作用。针对传统的GM(1,1)模型存在精度低的问题,结合连续9组时间序列货运量数据,通过二次函数对背景值重构,从而修正参数向量,构建改进的GM(1,1)预测模型。并利用2017~2019年货运量对模型预测精度进行检验。结果表明:最大相对误差≯1%,后验残差比C为0.18,小概率误差P为1,模型预测精度高。
Freight volume prediction plays an important role in transportation infrastructure construction and regional economic development,while the traditional GM(1,1)model has the problem of low accuracy.Combined with nine consecutive groups of time series freight volume data,this paper uses quadratic function to reconstruct the background value,so as to modify the parameter vector and construct an improved GM(1,1)prediction model.The prediction accuracy of the model is tested by using the freight volume from 2017 to 2019.The results show that the prediction accuracy of the model is high,the maximum relative error is≯1%,the posterior residual ratio C is 0.18,and the small probability error P is 1.
作者
袁志兵
YUAN Zhi-bing(Yiwu Industrial & Commercial College,Jinhua,Zhejiang,322000,China)
出处
《建材技术与应用》
2021年第5期23-26,共4页
Research and Application of Building Materials
基金
2021义乌工商职业技术学院科研项目(YB2021WY613-02)。
关键词
道路运输
货运量预测
改进GM(1
1)
背景值
road transportation
freight volume prediction
improved GM(1,1)
background value