The Research Of Network Queuing System Under Bounded Rationality Condition

The customers in the limited resources stochastic service system such as the World Expo, because the impact of its existing service preferences, the non-optimal transfer queues decisions made under the limitations of their computing ability and knowledge, and impatience emotion, resulting they could only be bounded rational manager rather than fully rational Economic Man. These characteristics of customers, making the queuing system is under the influence of bounded rationality environment, at last leading to the rules of the queue system being complicated. If such a bounded rational factor is not taken into account, then the result by analyzing such queuing system will be far different from the actual processing. At the same time, taking into account the need for such a queuing system is the system in the response focused on the impact peak situation, which led to the focus of our research should be the system transient performance index, rather than focusing on classical queuing theory steady-state properties. These factors make it difficult to analyze and process such queuing systems by classical queuing theory, so it is great significance for finding other ideas to analyze and process such queuing systems.This thesis first introduces the basic knowledge related to research needs, and then establishes queuing networks model under bounded rationality and considering impact peak condition which based on the characteristics of queuing system of we are studying and basic knowledge of queuing theory. After that, we sample and statistically process the actual data, and then research the distributions of customer arrival and service time, draw relevant conclusion. Based on these conclusions, we establish queuing network model example under bounded rationality and considering impact peak condition. Based on the model instance, we start the queuing system simulation. According to the analysis of simulation results, at last we obtained system optimization decision-making solutions and suggestions of the modeling instance object.