旅游客流预测模型的比较及其实证研究——以黄山风景区为例  

Comparison and Demonstration Studies on Tourism Passenger Forecast Models——take Huangshan as an example

作　　者：汪祖丞[1] 刘玲[1] 

机构地区：[1]华东师范大学资源与环境科学学院,上海200062

出　　处：《安徽师范大学学报（自然科学版）》2010年第3期286-290,共5页Journal of Anhui Normal University(Natural Science)

基　　金：世界遗产地保护专项资金项目(48054660)

摘　　要：本文以黄山风景区1979-2004年客流量为例,定量分析线性回归模型、指数平滑模型、灰色预测模型以及BP网络预测模型的应用及其差异问题,得出以下结论:1)如果序列较长(20a左右),可选择指数平滑预测模型;2)如果序列较短(7a以下),可选择线性回归预测模型;3)介于两者之间的可以考虑使用灰色预测模型;4)BP模型预测(拟合)值比较准确,但结构复杂.由于选择的模式不同,预测值也会有差别,需要根据试算结果选择适当的网络模式和函数参数.Based on the tourist volume in the Yellow Mountain between 1979 and 2004,this paper made the quantitative analysis and comparison of the application and differences of the Linear Regression Model,the Exponential Smoothing Model,the Grey Forecasting Model and the Forecasting Model of Artificial Neural Network.Following conclusions were made： 1） The Exponential Smoothing Model is suitable for long series which span about 20 years.2） The Linear Regression Model is suitable for short series that span less than 7 years.3） The Grey Model of GM（1,1） is suitable for medium series which span more than 7 years but less than 20 years.4） Predicted value of BP-ANN is more precise due to its complex structure.According to the differences among the selected models,predicted value will be changed.While there are morel interference factors,network modle and function parameters should be selected based on the trial calculation result.

关 键 词：客流量 预测模型 预测误差 

分 类 号：F590[经济管理—旅游管理]