基于Agent的景区游客行为仿真建模与应用——以颐和园为例

我国旅游景区内部客流管理一直以来都没有得到应有的重视。通过对颐和园景区"五一"小长假和"十一"黄金周客流持续2年的实际观测,采用基于Agent的仿真建模方法,构建了游客到达、游客移动、游客停留等景区游客基本行为模型及其仿真运行环境——基于Agent的景区游客行为仿真系统。系统的正确性验证表明,仿真系统的输出数据能够反映景区各停留点游客数量的真实情况;系统的灵敏性验证表明,入口游客数、游客在景点的停留时间以及流向景点方向的客流量3个参数能够引起仿真系统各景点游客数量的灵敏变化。所建系统能够预测景区客流的时空分布,能够通过变换仿真参数值进行各种客流调控措施的仿真实验,并给出具体调控参数以支持景区管理者的客流管理实践。

Agent-based Simulation Modeling of Visitor Behaviors at Tourism Attractions: The Case of the Summer Palace,Beijing

Visitor flow management in tourism attractions has not been given adequate attention in China. In recent years, as China＇s tourism has developed rapidly, visitor overload becomes more and more serious in key tourism sites. According to the statistical data on large scenic regions, in recent years, especially famous natural and historical tourism attractions have witnessed constant growth in terms of visitor quantity. The negative effects between the interaction of visitor behaviors and the tourism environment is becoming more serious, so are the issues regarding recreational quality and visitor safety. The scenic regions are desperate for effective solutions for the balancing problems of visitor use, protection and recreational quality. Computer simulation modeling of visitor behaviors in tourism attractions applies this technology to study how recreation use and resources inter-function and inter-infect basing on visitors＇ recreational behaviors. This will provide some information to help solve the balanced development problems between recreation use and resource protection. Wherein, agentbased simulation modeling is a good method to model visitor- centered, nonlinear, concurrency recreational systems that traditional data analysis models cannot describe. It is considered to be an important research direction in the simulation technology application in China tourism research and can adapt to the dynamism of the China recreation system. This paper proposes an agent-based simulation modeling approach to provide decision- making data for regulation of visitor flows in tourism attractions. By collecting and analyzing 2-year visitor flow data in China＇s short holidays of＂May 1st＂and＂Golden Week＂of＂October 1st＂in the Summer Palace, visitor basic behavior models of arrivals,movement and stay are proposed, including a Poisson distribution model for visitors entering tourism attractions, a Markov decision model for choice of scenic spots among tourism attractions, and an Erlang and exponential distribut