search

LyapunovJingci / Warehouse_Robot_Path_Planning

A multi agent path planning solution under a warehouse scenario using Q learning and transfer learning.🤖️
MIT License
119 stars 22 forks source link

how to run the code ? #1

Closed henryliuliuliu closed 3 years ago

LyapunovJingci commented 4 years ago

Hi! Thank you for your interest to this project. After cloning this project, please first change line 45 in warehouse_test.py and line 17 in RLBrain.py to your own file path. Then run file warehouse_test.py to see the simulation.

henryliuliuliu commented 4 years ago

I cloned your new code, and I found the same problem as follows： File "c:\Users\efeff\Desktop\Warehouse_Robot_Path_Planning-master\RL_brain_Robo1.py", line 23, in init f = open(RLBrain.FILEPATH + '\q_table1.txt', 'rb') I cannot found this file in your repository, where is the q_table1.txt?

LyapunovJingci commented 4 years ago

My bad. I just upload them. Thank you!

henryliuliuliu commented 4 years ago

Thank you! one more question, if I want to retrain the Q-table, should I change the currentstate to state0 and then state1 so as to state2?

LyapunovJingci commented 4 years ago

State 0 corresponds to training the map State 1 corresponds to training the robots to return to the starting point State 2 corresponds to training the robots to reach the target These three states follow the three training steps in the flow chart in Readme. Thanks!

henryliuliuliu commented 4 years ago

I'm sorry, I am a little puzzled, then how to test the trained network?