scheduling

[diana4]

[ FlexSim 20.1.3 ]

1115.fsm how can i dispatch this model usingshortest processing time and longest processing time,ERD(FCFS, ready time), ECT(ready time+processing time),Because my three products will be reflowed, for example, a machine will first process item1=0, and then reflow process item1=2, so how can the machine be ordered by processing time ASC before selecting item1=1 products in the SPT setting? I already have the processing time of each machine for three products in the global table,

[joerg_vogel_HsH]

You can read table values in function for Process Time. You can assign label values at your items entering your processors. You read such labels to identify table row and column to get table cell value as your process time. If you compute your process time by different table cells you need more label values to get evaluated.

[diana4]

hi @Joerg Vogel thanks your answers, I have setting labels and identify table row and column,but i don't know how to set the setting in SPT and anthor rules, Because my three products will be reflowed, for example, a machine will first process item1=0, and then reflow process item1=2 ,I want to Order by processtime ASC first ,and then select item=2 ,how to write the code in Query of the Pull Strategy?

[jason_lightfoot_adsk]

Put the expressions for ERD, ECT, SPT on the lists an when you pull you can say ORDER BY ..... ASC etc. Use the value and puller keywords in the expressions to refer to the jobs and resources (or vice versa) to determine the ERD and ECT based on the jobs that you've already assigned to that resource. I'd keep a track of those use a resource label with array or map data.

Also - your model looks like it should be using Object Process Flows since there is much duplication.

[moehlmann_fe]

Example model:

The two labels "NextProcessingTime" and "TotalPendingTime" are added in the On Creation trigger of the source, in addition to the "Type" and "Step" label.

The queues after the source push the items onto a list (onto different partitions, so that the two clusters can't access each others items). The labels are added to the list as fields. This allows to filter and order the items by the label values when pulling them.

(Only Type 1 and 2 can be pulled, ordered by the label "NextProcessingTime")

In the upper object cluster, the processors pull by the "NextProcessingTime" in ascending order, in the lower one by "TotalPendingTime". After running the model for some time, you will see that in the upper queue mainly accumulates green items at step3, since that step has the single longest processing time. In the lower queue fresh red items accumulate, since they have the overall longest processing time.

job-ordering.fsm

[a9080109]

The problem now is that I have a lot of q in the same place (because the products entered are different products). If so, is there a way to sort them? Or I can combine q into one. I would like to ask how to rewrite it.

[moehlmann_fe]

As long as the items are distinguishable by a common label you can place them all in a single queue and filter for the correct ones in the query. See my attached model above in which three different types all share a queue.

[a9080109]

I don't understand why my model isn't pushed into the list1115-+柔.fsm

And I have to make sure that my itemN=M is a must

I have two products now, one is blue, item1=0, but the spt will be 60, brown is item=4, but the spt will be 50, and under normal circumstances, it will be an object with a smaller spt.

The yellow part will be item1=3 I have to make sure the tags are correct as it will affect my order.

[a9080109]

@Joerg Vogel @Jason Lightfoot @Felix Möhlmann

[a9080109]

I want to ask this question, is there an answer?

[jason_lightfoot_adsk]

I think you have asked this elsewhere - in particular we have answered it again here.

Additionally, with the aims of transparency, academic integrity and giving local distributors the chance to support students, we prefer that student posts are not created as or altered to be private posts.