Cancel
Showing results for 
Search instead for 
Did you mean: 
Highlighted

Different arrival patterns on different days

Experimenter
Experimenter

Hello,

 

I have a problem with setting the arrival of the patients in the hospital.

On the working days of the doctors the arrivalrate of the patients is higher than other days, besides the doctors have different working schedules.

 

I tried with the ShiftCalendar to let patients arrive on different days, but when I run the model it breaks down at the moment it calls the calendar.

 

Somebody knows how to solve this?

 

-- Program in attachments, problems occur at .models.hospital.Endocrinology.scheduling.Shiftcalendar

3 REPLIES 3

Re: Different arrival patterns on different days

Solution Partner Phenom Solution Partner Phenom
Solution Partner Phenom

Hello Elieke,

 

There are some Source object with an awfully small lambda parameter for the Poisson distribution. I suspect this causes some of the problems; when I change the parameters to 1 second, it does not hang anymore. Later on you will find a possible infinite loop in the method "Schedule", line 66.

 

Looking at your model, it seems rather complicated for what it should do. For instance, the Drain object will already gather a lot of statistical information on your MUs (based on the object name) under Type statistics. This may already be enough.

 

And I noted that you do not clean up tables in a reset. Maybe this also helps.


Gert Nomden
Senior Consultant | Tecnomatix Product Manager
cardsplmsolutions.com

Re: Different arrival patterns on different days

Experimenter
Experimenter

Hello Gert,

 

Thanks for the fast reply.

 

As good as I know, the lambda parameter is the amount of arrivals per second, isn't it?

 

In mij case, for example the amount of Telephone calls on a working day for Doctor A is 4. A working day contains 28800 seconds, thus the number of arrivals per second is 0,0014. What is wrong in my thinking? As well, if I change the Poisson distributions to exponential distributions (in case of the example 02:00:35 as far as I know), the model still breaks down at the ShiftCalendar called WorkingDays.

 

I hope you can help me.

 

Elieke van Sark

 

 

Re: Different arrival patterns on different days

Solution Partner Phenom Solution Partner Phenom
Solution Partner Phenom

Hi Elieke,

 

The Interval-parameters determines the inter-arrival time (not # arrivals). So, without further checking, I think the negexp is a better choice. With your current Poisson settings, you get many zero interarrival times (since the result of the function is 0, 1, 2 ... seconds).

 

If I change the Interval parameters in ALL Source objects, then the model does not hang.


Gert Nomden
Senior Consultant | Tecnomatix Product Manager
cardsplmsolutions.com