logic combinner
Hello sir, I might need your help!
The problem I just mentioned is quite similar to the one described in the linked post, but I’m having trouble integrating both approaches.
In the linked example, the logic is implemented using nextcp for looping through control points. However, in my model, some control points require a stop while others do not, so I can’t fully adopt that method.
My current idea is to use groups to manage this behavior, but it’s not going smoothly.
Also, I noticed that in the RGV task, there’s a “Travel to Next CP” action, and I’m wondering if this could be a potential breakthrough point — though I still don’t have a clear idea yet.
Finally, my task assignment is designed to push tasks into a centralized task list table, which is quite close to what I want to achieve.
I’m wondering if it’s possible to add some conditions within the current setup so that tasks are only assigned to the AGV that actually reaches the designated point.
I tried to change some of it. I have tried my best. Please guide me again. Thank you.
It's quite hard to follow the logic in the model since I don't understand any of the chinese activity names. But I can try to give some general advice how the model could changed.
Tt seems that the model works by tokens first pulling an AGV (purple activities), then push themselves to the AGV task list so that the previously pulled AGV can start the task the next time it arrives at a control point.
What you can do is remove the purple activities and push the token to the task list without knowing which AGV will do the task. Then add a field to the task list that reads which CP is connected to the object that currently holds the item (this assumes that such a connection always exists).
An AGV token can then only pull tasks at the control point it just arrived at (or one connected with a work forwarding connection). The maximum wait time (meaning the delay time for which the AGV waits for a task) could be given by a label on the CPs. So overall this wouldn't use my AGV logic at all and instead just change when/how an AGV is assigned to a task.
I don't understand what you mean with "if the workcp in the WorkForwarding List isn't pulled out, then the next point will need to wait 5 seconds".
Tasks are pulled in the activity marked in yellow. Wait times for tasks would best be implemented as a maximum wait time in there. If a task is pulled the token continues in the same block of activities (green arrow). If no task is pulled, the token is released to the top right (red arrow). Any delays depending on a task being pulled or not would thus be implemented at those points.
What I want to understand is the model mentioned in the link! Like the picture, if the workcp is not occupied, the car will stay for 5 seconds. How can I apply this logic to my model?
First you need a way to reference the "workcp" based on the control point the AGV arrives at. In the other model this is the purpose of the WorkForwarding connection. In this model those connections are drawn in the opposite direction though (from workcp to "waitcp"). Many of the pairs also have a purple "Break" connection going in the opposite direction but not all. I am going to assume that all stations where this wait time should be applied will have those connections.
Then you can check if the workcp is allocated like in the other model and determine the maximum wait time based on this.
I think this combination is very successful, but after observation, I found that some indicators of my model are 0? Is there something I did not do well? Or is there something else that caused it to become 0?
I can't help you there much. All I can say is that this performance meausure shows the throughput of the activity marked in red in the screenshot below. I don't now if/when/why tokens are supposed to move through that section.
Why do I need to add extra points like this? Because I need to assign tasks only when the car hits the CP point, but my detour path does not necessarily reach the point. So I want to judge at some points whether there are tasks at these points. Do I need to send the car there to accept the task?
The "connectedCP" list field and query is currently set up to expect a single CP. You could change this though so the field instead stores an array of all CPs connected via "point" connections. The query would then be changed to check if the current CP of the AGV is part of that array.
The first error I'm getting when I run the model is caused by a CP not having a NextWorkPoint connection.
Sir, I have connected to NextWorkPoint but the error still occurs at the same time. I would like to know where I went wrong. Thanks.
There might be other CPs that are missing a connection as well. When an error occurs, put a stop time at the time of the error and check the token moving through the activity mentioned in the error message. Check if all labels are make sense and if the "currentcp" has a NextWorkPoint connection. The code for the "nextcp" label in the marked activity seems to be responsible for most errors. I edited the expression for the "currentcp" label value because the previous code didn't make much sense (was needlessly complex).
In the "connectedCP" field you are pushing all controls points connected as a "Location" to the array. These are not part of the NextWorkPoint loop though. The queue the item is in when pushed to the AGV work list (I assume its the one marked with green in the screenshot below) needs to a have connection (for example the "point" connection you had earlier) to a CP that is part of the NextWorkPoint loop and ahead of the spur path (blue). This connection then needs to be used in the "connectedCP" field instead of Location. That way, an AGV that is arriving at the blue CP can pull the item/task.
Hi sir, my model doesn't have any major errors right now, but what I want to do might be a little different. When I touch a point, I should be assigned a task to pick up the goods at the location; otherwise, it should just pick up the goods at the point, which isn't what I want.
I am sorry, but your sentence doesn't make much sense to me. What is the red arrow and 15 supposed to show in the screenshot?
When a token is pushed to the AGV task list, the "connectedCP" field gets all CPs that are connected to the object the item is in with a "point" connection. When an AGV arrives at one of those points it can pull the task token. How long it waits at a CP still functions like in the example model I provided in another post. If the CP the AGV arrived at is connected to a different CP with a "Workforwarding" connection (the connection must originate at the CP the AGV is at), a check is made if the connected CP is allocated or not. If not, the wait time is 5s. If yes, or no such connection exists, the wait time is 5s.
