dimitrirenard / from-reinforcement-learning-to-reality-generating-structured-text-logic-controller-

Proof of concept of an implementation of control tables learn with Reinforcement Learning in a PLC

From-Reinforcement-Learning-to-Reality-Generating-Structured-Text-Logic-Controller-

This repository contains a PLC implementation of the proposed methodology:

• TIA Portal Projet
• Control table learned by Reinforcement Learning
• Operating part simulation Factory IO scene

Task	Component	Description	Non-parallel compatible tasks
T1	Feeding Conveyor 1	Feed part to position 1	T12
T2	Feeding Conveyor 2	Feed part to position 3	T13
T3	Robot	Rotation without part 0 to 1	T6
T4	Robot	Rotation with part 1 to 2	T12
T5	Robot	Deposit part	T8, T9, T10, T11
T6	Robot	Rotation without part 0 to 3	T3
T7	Robot	Rotation with part 3 to 2	T8, T13
T8	Robot	Rotation without part 2 to 0	T5, T7, T10, T11
T9	Output Conveyor	Evacuate part	T5
T10	Robot	Rotation without part 2 to 1	T5, T8, T11
T11	Robot	Rotation without part 2 to 3	T5, T8, T10
T12	Robot	Take a part in position 1	T1, T4, T8
T13	Robot	Take a part in position 3	T2, T7, T8