|
|
Visitors arrive at the ground floor according to a poisson process, on average one visitor per minute. About 75% of the visitors like to go to one of the upper floors; the other visitors have to do some obscure business in the basement. Because these visitors are a little bit lazy, they certainly do not take the stairs, but they all wait for the elevator to travel to one of the upper floors or to the basement. There is only one elevator with a carrying capacity of 6 people.
Construct a simulation model for this situation, using only modules from the Basic Process panel. Because we don't know what is happening at the other floors (we even don't know how many floors there are), only the ground floor is modelled, making abstraction of the upper and lower floors. Assume the following policy: visitors are polite and wait in FIFO order, but if the elevator is coming down from the upper floors, and visitors are waiting to go to the basement, then they get priority (in other words: the elevator keeps on going down, and visitors for one of the upper floors do not jump in an elevator going to the basement); vice versa, if the elevator is returning from the basement and visitors are waiting to go to one of the upper floors, then they get priority (in other words: the elevator keeps on going up, and visitors for the basement do not jump in an elevator going up). Assume that an empty elevator automatically returns to the ground floor.
A normal person needs about one second to enter the elevator, but some people need up to five seconds. No one can enter the elevator in less than 0.5 seconds.
Measurements of service times are available and are given in files ElevatorTimesUp and ElevatorTimesDown (.TXT format): ElevatorTimesUp contains the times measured from the moment the elevator going up leaves the ground floor, until the elevator returns at the ground floor and is emptied; similarly ElevatorTimesDown contains the times measured from the moment the elevator leaves the ground floor down to the basement, until the elevator returns at the ground floor and is emptied. Use the InputAnalyzer to find appropriate distributions for the service time of the elevator (if possible), and use these distributions in your model. Explain why a particular distribution is suggested by InputAnalyzer and/or selected.
Run the model for 100 hours (single replication) and generate the one-page SIMAN Summary Report (.OUT file).
During peak hours the arrival rate quadruples to four visitors per minute. The same policy holds as before, but during peak hours men are even more polite, and allow women to enter the elevator first. Again run the model for 100 hours (of course this peak is not going to last for 100 hours, but this is just to check whether your model can cope with this situation). Is there a significant difference in the behavior of the model?
What is your final judgement? Is this a good representation of reality? If not: what fundamental concepts are missing?
[ 本帖最后由 marygrace 于 2008-10-10 05:13 编辑 ] |
本帖子中包含更多资源
您需要 登录 才可以下载或查看，没有帐号？注册
x
评分
-
查看全部评分
|
[复制链接]
发表于 2008-10-10 05:12:26
