I want to make a time standard deviation of 1 ± 0.5 minute "when patients entering to service center take ticket and seat."
and also make another process time of 7 min with standard deviation of 2 min for a processor machine.
In your other question you stated:
They take 1 ± 0.5 minute uniformly to do so.
In that case you would use a uniform distribution:
1 + uniform(-0.5, 0.5)
@Mohammad reza A, you can get around negative values of a normal distribution, but that will falsify the distribution. There are two possible scenarios, you build a loop that works until the distribution comes up with a positive value or you cut the negative values and set them to 0.
double statDistro = normal(1,0.5,1);// negative values are possible
double timeValue = Math.max(statDistro,0);// negative values set to 0
Thank you @Jörg Vogel. should I define it for each processors? for example in one of the processors the visit lasts to be normally distributed with a mean of 7 minutes and a standard deviation of 2 minutes. should I make a process flow for each?
@Mohammad reza A. When using nomal distribution, the standard deviation time of 1 ± 0.5 minute means minus time in the case of 1 - 3* 0.5. The service time can not be minus time and so you meet error or the model can not run at any time.
You can check it and should fix the time or your model.
8 People are following this question.
