【正文】 ns for c), the objective function minz (c) = Csc + CwLs (c), which Ls for the number of stay (formula (4)), c can only take an integer, located c * Is to make the minimum objective function c from the point, c * meet z (c * 1) ≤ z (c *) = csc * + cwLs (c *) ≤ z (c * +1), Ls = Ls (c) Reduction in the Ls (c *)Ls (c * +1) ≤ cscw ≤ Ls (c * 1)Ls (c *) (5) Followed by puter simulation calculated Ls (1), Ls (2), Ls (3), … the difference between the two neighboring, which fell to see constant between the two to determine the cost to patients and the cost of hospital services and Desk achieve optimum number of the optimal solution c C *. On the issue of optimization services program When the patients arrived at an average rate of increase in services from which the average increase in strength captain L too, and even the service intensity ρ1 captain tended to unlimited, the average service rate unchanged on the case can only increase Desk. Below are two to discuss them and help desk services, the average rate equal to the situation. We can help desk was informed that two forms of service an average of two captain L, the waiting time for the ratio of W: 2L1L2 = W1W2 = 1 + ρ2 (ρ2 = λ2μ 1) On the people most concerned about in terms of waiting time of 1 Similarly can prove that: a number of parallel desk in the queuing system, lineline team than many teams tied for the program has significant advantages. For multiple attendants random process, if only the waiting time for patients should be considered in a row only team. 2 Case Study A hospital operating room to have random services, the Census and Statistics of the 100 h patient treatment and to plete the operation of data. ① calculated the corresponding number of indicators。 System in a few patients () wait a few patients (who) patient length of stay (h) queue waiting time (h) service intensity ρ = λ / μ = coefficient of patients loss of time the operating room the probability of idle time peak hours, the probability pn = ② puting services strength ρ = λ / c μ = 1 an increase in the scale of the operation theatre after the same number of indicators System in a few patients () wait a few patients (who) patient length of stay (h) queue waiting time (h) of the two operating theatres idle time only a probability of idle operating room the probability of patients p1 = .Without waiting for the probability of patients must wait for the probability of . Based on the above data available indicators: There is only one operating room department patients waiting for surgery time is times 84% of the operating room is the busiest time, only 16% of the free. If an operating room to be used to increase the probability is 42 percent, the probability is idle 58 percent, two operating theatres idle time the probability of , two operating theatres there is only one spare the probability of 34 percent. Based on the above data policy makers to decide whether an increase in the operating room, so as to provide managers decision support tools. 3 Conclusions To the hospital queue is a mon phenomenon, as patients and medical services at the time of the randomness, with the number of sources in theory is unlimited, and medical resources are limited, how the allocation of limited resources, the use of the line model theory And puter simulation, bined with the service record of relevant data, to make qualitative and quantitative indicators of the number, then the forecast, analysis and evaluation, by optimizing the design, implementation dynamic management, according to the hospital39。 the hospital has bee one of the core petitiveness, sustainable development into the hospital an important ponent part. With the continuous deepening of reform of the hospital, the hospital information to speed up construction will help hospital administrators grasp the general direction of reform, and strengthening the hospital management, improve the social and economic benefits.