question

Lorenzo_pio G avatar image
0 Likes"
Lorenzo_pio G asked Jeanette F commented

Application of reinforcement learning on a robot

1665234839424.png

Hi! That's my model. I wanto to apply Reinforcement Learning to the robot called Robot1. The idea is that the robot, based on the number of boxes on the pallet placed on combiner1, should choose on which combiner to deposit the box. Each pallet collects 10 boxes before passing the load to the left queue. Robot1 should help Robot2 load the boxes on the pallet present on combiner1 until the number of boxes on the pallet becomes more than 6. As soon as this happens, Robot1 must start carrying the boxes on the combiner called "Impacchettamento" First I want to identify as a variable the instantaneous number of boxes on the pallet present on combiner1. Next I want to identify as an integer the combiner that robot1 is using. I need these variables to identify the observations and actions of the RL.

Bakery1.fsm

FlexSim 22.1.3
robotreinforcement learningcombiners
1665234839424.png (225.8 KiB)
bakery1.fsm (422.9 KiB)
· 1
5 |100000

Up to 12 attachments (including images) can be used with a maximum of 23.8 MiB each and 47.7 MiB total.

1 Answer

Felix Möhlmann avatar image
0 Likes"
Felix Möhlmann answered Lorenzo_pio G commented

You should start by building a control logic that allows the two robots to act independently from one another, instead of just having them execute transport tasks generated by the queue. For example, you can have the queue push the items to a list and the use a process flow, in which tokens (that represent the robots) pull those items off the list. They can then decide where to unload them depending on whether there are pallets present on either combiner and how full those are.

This tutorial deals with task dispatching using lists.

The following expression would check if there is pallet on combiner1 and whether it holds more than 6 items.

  1. Model.find("Combiner1").subnodes.length > 0 && Model.find("Combiner1").first.subnodes.length > 6
· 46
5 |100000

Up to 12 attachments (including images) can be used with a maximum of 23.8 MiB each and 47.7 MiB total.