Hi everyone,
I’m working on a FlexSim RL project involving two machines that produce two product types with different task precedence relationships. The machines can fail in stages from healthy (stage 1) to fully broken down (stage 4). Maintenance can occur at any stage, with varying costs and times. My objective is to minimize both cycle time and maintenance costs.
Issues:
Slow Training: The training process is slow on my Intel Core i7 system (8 logical cores, 16 GB RAM). It doesn’t seem to fully utilize my CPU and memory.
Training Stuck: The training often stalls midway, possibly due to event-triggering issues in the FlexSim RL setup. I’ve tried different configurations without success.
Request:I’m looking for advice on optimizing system resource usage, best practices for RL event definitions in FlexSim, and strategies to prevent the training from stalling.
Any guidance would be greatly appreciated!
.fsm
MMALBPAUG19_Maintenance.fsm
flexsim_env
flexsim_env.txt
flexsim_training
