Possiblity excel to ED

All topics specific to modeling questions in Enterprise Dynamics
marlies
Posts: 301
Joined: Monday 17 January, 2011 - 09:28

Re: Possiblity excel to ED

Post by marlies »

Hi Ditmar,

1. Yes, when you change the growth factor it will be taken into account. Please have a look at the code that I added in the following location: atom arrivallist > tab arrivals then edit and field quantity has this code:

Code: Select all

cell(1,2,refGrow)*Max(0,Normal(93.8,41.75))
In this code you find the cell reference to row 1, column 2 in refGrow. So, the arrivallist reads the growth factor from the table. When you change it in the table it will be taken into account.
NB: I added the Normal distribution for the number of arrivals with an average of 93.8, you'll have to check if that's indeed the number you need.

2. The multiservice is the atom where trucks will "wait" until arrival. The code reads in which hour the truck will arrive label([HourArrival],i) and will pick a random time within that hour. So, it depends on what you need. I made the option that each hour of the day a certain % of trucks will arrive. If that's what you need, just leave the code as it is. If it's not what you want, please explain what you like to do. Note that you can change the % of trucks per hour of the day in the empirical distribution Arrival Hour.

3. Your description of the model looks good. If you are sure that 15 is enough for the queue capacity you can do that, if it's something you need to find out you better increase this capacity. For the explanation of the autofit tool go to Help and in the index search for autofit tool.

4. I am not sure what is happening in your model, it would help if you could attach it, so we could check.

Regards, Marlies
ditmar
Posts: 19
Joined: Friday 21 March, 2014 - 11:45

Re: Possiblity excel to ED

Post by ditmar »

Hi Marlies,

I did everything as suggested by you. Now i've made the complete model with also outgoing traffic which passes the same server because it also have to be weighted.
I did the following things:

- made an outging arrival list with the same label hour code.
- made a multiservice with an emperical distribution from outgoing traffic data.
-made a outgoing queue with the same waiting capacitiy
- Set the server on round robin, because protocol is: 1 in out, 1 out.
- Made a emperical distribution for the server

Now the following problems occur:
- I don't know if I used to autofit option in a proper way, because its still not very clear to me. You can't select an atom and then use autofit or something. So i just tried with all empirical distributions ( arrival and depature multiservice and the wheiging bridge) ? maybe you could check if right ?
- When I do a experiment with 30% increasing growth It says: 9 trucks average in queue and 2.48 hrs average waiting time for incoming and for outgoing a little less. Problem is 9 trucks seems okay but 2.48 hrs is very long. With experience i know it only can take 1 to 1.5 hrs max. And when I don't use growth I put in both tabels; 1 in stead of 1.3 it has the same outcome in and outgoing ? very strange.

I attached my model maybe you can have a look on any errors on my behalf.

Thank you very much for your assistence.

Best regards

Ditmar Rijk
-
ditmar
Posts: 19
Joined: Friday 21 March, 2014 - 11:45

Re: Possiblity excel to ED

Post by ditmar »

Hi Marlies,

When you made the empirical distribution for my model arrival list. Could you explain how you came by the distribution figures (max(0,Normal(93.8,41.7) because when i upload my figures I get a Neg.exponential(894.92) with very high score(28461). ?

Also when you upload the data with ActiveExcelX or a table do these Atoms need to stay in the model, so they can constantly read them ? Or is it sufficient to do it once and autofit the distribution and delete the tabel again ?

I uploaded my model for your information. As you can see the arrivals are still configured like you did and I made and Excelactive X connection to read the Server(wheigingbridge) times and translated that in to a distribution. The only problem is that my programm version of ED can only use one Excel Atom at the time and it seems that I'm doing something wrong with the depaturelist as the score of the autofit is much to high, and I guess the incoming data and outgoing are distributed the same?

Thankyou in advance,

Regards

Ditmar
outgoing + wheiging bridge data.xlsx
(695.09 KiB) Downloaded 253 times
Attachments
huidige situatie in & uit - met active excel.mod
(41.03 KiB) Downloaded 257 times
ditmar
Posts: 19
Joined: Friday 21 March, 2014 - 11:45

Re: Possiblity excel to ED

Post by ditmar »

huidige situatie in & uit - met active excel.mod
(40.85 KiB) Downloaded 251 times
Hello,

I've already found out some things myself. Expect for the arrival distribution? (mentioned in previous post) However the model is not working properly. Latest runs show: 15 trucks average in both queues and an average of 2.48 hrs waiting time.

Maybe some one can check if I did everything right and what do I have to change to get the proper result.
In reality it's most of the time 5-6 average in queue and maybe maximum 1 hour waiting time.

I think there is something wrong with impirical distributions and maybe some 4D scripts.

Find my model attached , the excel model was attached in the previous post.

Best regards

Ditmar
Post Reply