Flexsim job shop problem

[a9080109]

[ FlexSim 22.0.16 ]

I'm currently building a model for gig production and I'm not sure why my model is having issues. But what I can provide is some preliminary information on modeling.

1. I have 20 orders but my production order is randomly generated.

2. My processing time wants to be set to evenly distribute 1,100, but I am wondering if I can give him a processing time label at the beginning, and then the machine will process according to the processing time label on the item.

But now it seems that my model is not processed as I thought.

random.fsm

[joerg_vogel_HsH]

@mark zhen, what do understand under „evenly distributed“?

[a9080109]

The frequency of occurrence on a scale of 1-100 is the average occurrence

[a9080109]

@Felix Möhlmann @Jason Lightfoot@Kavika F

Hello, I have updated my content, but there is still a problem with the processing order. I don’t know what the problem is.random.fsm

[moehlmann_fe · Accepted solution]

In both your models you use a label ("processtime" and "pt" respectively) to refer to the column from which the process time should be read. In both cases, the values of the labels point to wrong columns or aren't even in the range of the table.

In your first model you have "duniform()" expressions in the "StationAssignments" table. You replaced these with number data without switching the cell nodes out of 'flexscript-mode'. As as result, only 0s are returned when reading from the table. It's probably easiest to just create a new table alltogether.

[a9080109]

But I fixed the random seeds and the order of StationAssignments in the new model, and I still can't run the model smoothly. There will be a lot of products that are not labeled smoothly. I want to solve this problem. As for the processing time, my approach is to read a new table and the table content will have the corresponding TYPE and PROCESSTIME. I think this approach should not be done. There will be problems.

[a9080109]

In both models, you use labels ("processtime" and "pt" respectively) to reference the fields from which the processing time should be read. In both cases, the label's value points to the wrong column, or isn't even within the scope of the table. (I don't understand where I made this mistake, the settings in the second model)

[moehlmann_fe]

When an item is created, a value of 1 is assigned to the "pt" label in Process Flow.

Processor1 uses this label to refer to the table column from which the process time should be read. Column 1 in that table is called "Arrival Time" and only contains 0s though.

I fixed this and the table issue in the attached model and reworked the processor connections.

random_2.fsm

One issue that remains is that the "date" label is missing on the items which the sink tries to read.

[a9080109]

But I want to know why there is a problem with the form? Can you explain to me the situation where I can't label it smoothly?

[a9080109]

So if the content of the table I want to read is an allocated form, how can I execute it smoothly?random-2.fsm

[moehlmann_fe]

The table cells in the "StationAssignments" table were set to contain FlexScript data. This only works if the entered data is in string (text) format.

You then apparently pasted the content of a different table in, overwriting the data with numeric data. But because the nodes are still compiled as FlexScript, every cell in the table only returned 0.

[a9080109]

So what changes did you make to my second model to correct my problem? The table content of my second model should be a number? It should run smoothly, right?

[moehlmann_fe]

I changed the cells to "Number Data" and I removed the output queues that received materials from the processors. Some were missing a connection back to one of the processors. Instead the items re-enter Queue3 if they are not done yet.

As I said, apart from the issue with the missing "date" label when entering the sink, the model seemed to run fine.

[a9080109]

@Felix Möhlmann so i want to use the duniform to set date label but its doesnt work!!

[a9080109]

random-2 (1).fsm

[moehlmann_fe]

FlexScript doesn't work in the label columns of the source schedule.

Move the distributions to a separate table, whose column is set to FlexScript and assign the values in a trigger of the source.

Also, the expression has to either be

return duniform(...);

or

duniform(...)

Either "return" and semicolon or neither.

random-2-2.fsm

[a9080109]

@Felix Möhlmann

I now find that there is a problem that when pulling my model, I should give priority to the things that have been processed for production, but my model does not seem to do this. This means that the artifacts in step 2 should take precedence over step 1 and so on. In addition, I would like to ask what is the function of my pull requirement?

[moehlmann_fe]

Then pull the item with the highest "step" value instead of ordering them by whether their type is equal to the respective action parameter.

The Pull Requirement filters which item can be pulled. If it returns 0 (false) the item can not be pulled. Any other numerical value (true) means it can.

The one in the model only allows the processor to pull items whose next step is assigned to it in the "StationAssignments" table.

[a9080109]

I don’t quite understand what you mean, but what I want to express is that the priority of processed objects should be greater than that of unprocessed objects. Have I successfully implemented this function?

[a9080109]

I want to know why I can't execute spt. I confirmed that my list has successfully recorded my tags.spt0501.fsm