Reinforcement learning question

Question

question

Luis Osvaldo Gaytan asked Jun 28, '24 Felix Möhlmann answered Jul 1, '24

Reinforcement learning question

Hello mates, I tried to use the reinforcement learning feature to determine how many totes we need to use to maximize the throughput, this exercise should determine the number of the final product with 2 totes it's the same with 3 or 4 totes but less with only 1 tote. When I tried to use the python make show this error and stop working.

Best regards!

RL_Tote.zip

Software Version:

FlexSim 24.1.0

reinforcement learning

image.png (16.6 KiB)

image-1.png (143.1 KiB)

rl-tote.zip (41.8 KiB)

Answer 1 · 2024-07-01T06:41:01Z

Felix Möhlmann answered Jul 1, '24

The Reinforcement Learning tool is not set up to do anything in the model you uploaded. Did you upload an older version of the model, because running the algorithm with the uploaded model should not yield anything.

Using RL also really doesn't make much sense here. RL can be useful if complex decisions need to be made during the runtime of the model based on the current state of the system. The RL algorithm would likely figure out which number of totes is the best if a dummy event at the end of the simulation gives it a reward based on how well the system performed.

But the same can be done much quicker and easier by defining a useful performance meausure (which you already have) and running each possible scenario in the experimenter.

1719816041304.png (30.0 KiB)

question