由于信息稀缺性、服务系统的不确定性以及顾客认知能力的有限性, 顾客很难准确估计接受服务后的预期效用, 进而做出理性决策。本文利用Logit选择模型描述有限理性顾客的策略行为, 基于排队博弈理论, 在几乎不可见和完全不可见两种信息水平下分析多重休假M/M/1排队系统中的顾客决策、机构利润与社会福利, 并通过数值算例对比不同信息水平下顾客有限理性水平对机构最优利润和社会福利的影响。结果表明, 在两种信息水平下, 有限理性顾客的均衡策略存在且唯一; 存在唯一最优价格, 使得机构利润最大化; 存在唯一最优价格, 使得社会福利最大化; 随着平均休假时间缩短, 顾客更愿意加入系统, 机构可以获得更多利润, 社会福利也会提高。此外, 在完全不可见情况下, 社会福利关于顾客有限理性水平的变化规律依赖服务价格, 是顾客有限理性水平的单峰函数; 存在一个特定有限理性水平, 使得机构利润和社会福利同时最大化。

Due to the scarcity of information, the uncertainty of the service system and the limited cognitive ability of customers, it is difficult for customers to estimate the expected utility after receiving the service accurately, and then make rational decisions. In this paper, the Logit choice model is used to describe the strategic behavior of boundedly rational customers, based on queueing game theory, the customer equilibrium strategy, revenue and social welfare is analyzed under almost unobservable case and fully unobservable case, the influence of customer boundedly rationality on the optimal revenue and optimal social welfare under the two information levels is compared through numerical case. The results show that, under the two information levels, the equilibrium strategy of boundedly rational customers exists and is unique; There is a unique optimal price that maximizes the revenue; There is a unique optimal price that maximizes the social welfare; As the average vacation time decreases, customers prefer to enter the system, revenue increases and so does social welfare. In addition, under the fully unobservable case, the variation of social welfare with respect to the bounded rationality level of customers depends on the service price and is a unimodal function of the bounded rationality level of customers; There exists a specific bounded rationality level such that revenue and social welfare are simultaneously maximized.