基于SEA的AGA-SVR节假日客流量预测方法研究  

A Study on the Forecasting Method of AGA-SVR Modeled Holiday Tourist Flows Based on SEA

作　　者：陈荣[1] 梁昌勇[2] 葛立新[3] CHEN Rong LIANG Changyong GE Lixin(Department of Economic and Management, Bengbu University, Bengbu, Anhui 233000, China School of Management, HeFei University of Technology, HeFei 230009, China Department of science, Bengbu University, Bengbu, Anhui 233000, China)

机构地区：[1]蚌埠学院经济与管理学院,安徽蚌埠233030 [2]合肥工业大学管理学院,安徽合肥230009 [3]蚌埠学院理学院,安徽蚌埠233030

出　　处：《旅游科学》2016年第5期12-23,共12页Tourism Science

基　　金：国家自然科学基金重点项目"基于云的管理信息系统再造研究"(71331002);国家自然科学基金面上项目"基于行为决策的隐性目标决策模型与方法研究"(71271072);安徽高校自然科学研究重点项目"智慧景区背景下的旅游客流量预测方法研究"(KJ2015A143);安徽省教育厅2016年高校优秀拔尖人才培育资助项目"高校优秀青年骨干人才国内外访学研修重点项目"(gxfx ZD2016283)

摘　　要：节假日旅游客流量由于游客在短期内大量集中于同一地点、不同节假日休假时间长短、所处季节等不同,呈现复杂非线性特征和明显季节性特点。本文提出一种基于季节指数调整（Seasonal Exponential Adjustment,SEA）的自适应遗传算法（Adaptive Genetic Algorithm,AGA）-支持向量回归（Support Vector Regression,SVR）预测模型,即基于SEA的AGA-SVR模型,并用国内著名5A级风景区黄山2008~2012年节假日客流量数据对模型进行验证。研究结果表明,基于SEA的AGA-SVR预测模型能够准确处理节假日旅游客流量预测中的非线性和季节性问题,较AGA-SVR和GASVR等方法具有更高的预测精度,在旅游预测领域应用前景广阔。In virtue of tourists amassing at the same scenic spots in a short term and the different lengths of holidays and seasons, holiday tourist flow is found to present complicated nonlinear features and obvious seasonal characteristics. This paper proposed a forecasting model which combined seasonal exponential adjustment with adaptive genetic algorithm （AGA） and support vector regression （SVR）, and used the new holiday daily data from the famous 5A scenic spot Mount. Huangshan during 2008 - 2012 to verify its effectiveness. The results indicate that the model, superior to AGA-SVR and GA- SVR, is an effective way to forecast daily holiday tourist flow.

关 键 词：季节指数调整 支持向量回归 自适应遗传算法 节假日旅游客流量预测 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]