<![CDATA[Discrete event simulation is widely used to evaluate the performance of a queuing system. In those simulations, the data distribution and parameters are determined to approximate the real conditions of the processes that occur in a queuing system. However, the queuing systems with their data distribution and parameters have so far been very specific to a particular system, not general to other systems that run a similar service and own a similar system structure, for example puskesmas, B and C. This research analyzes how much differences the queuing system produces when the system uses different parameters for their data distribution. The analysis is carried out by building the M/M/1 and M/M/2 queue models, then benchmarking several parameters of the exponential distribution and triangular distribution which are used to generate the time lag between arrivals of queue objects and service times. An increase of 60 in the exponential distribution parameter (1 minute patient arrival rate) causes the average number of patients in the queue to decrease by 4 people, and the frequency of large queues to decrease by 5 times. A difference of 0.5 in the triangular distribution parameter (faster service time 0.5 minutes) results in a reduction in waiting time of 0 - 0.2 minutes, and has an impact on the frequency of queues and large queues which can be reduced up to 9 times. Meanwhile, the difference in the number of servers which is one and two on the system can reduce the queue size from 16 to zero]]>
Copyrights © 2023
