Deep reinforcement learning enables conceptual design of processes for separating azeotropic mixtures without prior knowledge

Keyphrases

• Design Process 100%
• Azeotropic Mixture 100%
• Deep Reinforcement Learning (deep RL) 100%
• Without Prior Knowledge 100%
• Planning Problem 66%
• Flowsheet Optimization 66%
• Flowsheet 66%
• Reinforcement Learning 33%
• General Tasks 33%
• Search Space 33%
• Conceptual Approach 33%
• Reinforcement Learning Agent 33%
• Process Engineering 33%
• Continuous Parameters 33%
• Feed Composition 33%
• Chemical Engineering 33%
• Chemical System 33%
• Distillation Column 33%
• Distillation 33%
• Distillation Boundary 33%
• Composition Approach 33%
• Heteroazeotropic Distillation 33%
• Process Synthesis 33%
• Curved Boundary 33%
• Process Topology 33%
• Reinforcement Learning Approach 33%
• Multiple Chemicals 33%
• Discrete Process 33%
• Training Cycle 33%

Chemical Engineering

• Flowcharting 100%
• Reinforcement Learning 100%
• Process Engineering 25%
• Chemical Engineering 25%
• Process Synthesis 25%