基于季节SVR-PSO的旅游客流量预测模型研究  

Forecasting tourism flow based on seasonal PSO-SVR model

作　　者：陈荣[1,2,3] 梁昌勇[1,3] 陆文星[1,3] 宋国锋[3] 梁焱[4] 

机构地区：[1]合肥工业大学管理学院,合肥230009 [2]蚌埠学院经济与管理系,蚌埠233000 [3]过程优化与智能决策教育部重点实验室,合肥230009 [4]黄山风景区管理委员会,黄山242700

出　　处：《系统工程理论与实践》2014年第5期1290-1296,共7页Systems Engineering-Theory & Practice

基　　金：国家自然科学基金(71331002;71271072);安徽高校省级自然科学研究项目(KJ2012B097);安徽省科技厅科技计划项目(10120106011)

摘　　要：准确的旅游客流量预测对旅游风景区有着决定性的意义.受多种原因影响,旅游客流量预测不仅呈现复杂非线性特点,而且显示出典型的季节性趋势,尤其在旅游旺季.文章提出一种季节支持向量回归(seasonal support vector regression,SSVR)和粒子群算法(particle swarm optimization,PSO)结合模型,即SSVR-PSO,实现对旅游客流量的预测.来自国内著名5A级风景区黄山2008-2011年最新月客流量数据仿真结果显示,SSVR-PSO模型预测精度明显高于SVRPSO、SVR-GA、BPNN、ARIMA等方法,是进行旅游客流量预测的有效工具.Accurate tourism flow forecasting has always been the most important issues in the scenic spots. Due to various influences, tourism flow forecasting reveals not only a rather complex nonlinear characteristics but also seasonal trend, particularly during tourism peak periods, This article presents a tourism flow forecasting model that combines the seasonal support vector regression model with particle swarm optimization algorithm （SSVR-PSO）, to forecast tourism flow. Finally, the new monthly data sets from a famous 5A scenic spot Mount Huangshan during 2008-2011 are simulated repeatedly. Empirical results demonstrate that a SSVR-PSO approach is an effective way to forecast tourism flow, outperforming SVR-PSO, SVR-GA, BPNN and ARIMA.

关 键 词：支持向量回归 季节调整 粒子群算法 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]