基于多层步进动态预测的客源流向模型仿真  

Simulation on the Model Of Tourists Flow Direction Based on Multi-Layer Stepping Dynamic

作　　者：王新亚[1] 

机构地区：[1]平顶山教育学院旅游与酒店管理部,河南平顶山467000

出　　处：《计算机仿真》2014年第5期268-271,共4页Computer Simulation

摘　　要：研究旅游客源建模优化问题,由于客源流向的目的地带有倾向性、不可预估性和瞬时改变性,受到单个人员可以引导干扰和思想导向的影响较大,无法形成长期固定的导向特征;传统的客源流向估计预测方法采用模式匹配的非线性对应模型,一旦遇到频繁外部导向对客源的刻意干扰,简单的匹配模式陷入多匹配,盲匹配等问题,流向估计的准确度较低,对于客源的预测效果差;提出一种多层步进动态预测客源流向估计方法,采用将游客分层的方法设置多层,在多层分布的基础上,通过步进的方法实现系统动态预测,由于分层效果模型可以对游客进行详细的模型匹配描述,预测效果更佳,采用某景点的实际客流做实验分析,结果显示,采用基于多层步进动态预测的客源流向估计方法,客源流向的估计精度更高,接近实际的客源分布,准确度较高。The destination of tourist flow direction is tendentious, unpredictable and changeable, and can be af- fected by the interference of individual＇s guidance and ideological orientation. The long-term fixed oriented features are to be formed. This paper presented a new method to estimate tourist flow direction based on multi-layer stepping dynamic prediction theory. Visitors were set multi-layers and then the system dynamic prediction can be realized by multi-layer stepping method. Because the layered model can make model match description with detailed information in visitors, the predict effect is better. An actual tourist flow in one sight spot was analyzed. The results show that our new method can provide more accurate estimation of tourist flow and the gap between actual and estimated tourist dis- tribution is very small, the degree of accuracy is high.

关 键 词：多层步进 动态预测 客源流向 

分 类 号：TN391[电子电信—物理电子学]