en
TUM Campus Straubing
BIT - TUM Campus Straubing > Articles by: Dominik Grimm
New Publication: Automated Flowsheet Synthesis Using Hierarchical Reinforcement Learning

New Publication: Automated Flowsheet Synthesis Using Hierarchical Reinforcement Learning

Recently we showed that reinforcement learning can be used to automatically generate process flowsheets without heuristics or prior knowledge. For this purpose, SynGameZero, a novel two-player game has been developed. In this work we extend SynGameZero by structuring the agent’s actions in several hierarchy levels, which improves the approach in terms of scalability and allows the consideration of more sophisticated flowsheet problems. We successfully demonstrate the usability of our novel framework for the fully automated synthesis of an ethyl tert-butyl ether process.

New Funding: Glycoside Production

New Funding: Glycoside Production

We successfully attracted funding for a new project from the Bavarian Ministry of Economic Affairs, Regional Development and Energy. More details will follow soon on our project page.

New Publication: Automated Process Synthesis Using Reinforcement Learning @ ESCAPE31

New Publication: Automated Process Synthesis Using Reinforcement Learning @ ESCAPE31

A novel method for automated flowsheet synthesis based on reinforcement learning (RL) is presented. Using the interaction with a process simulator as the learning environment, an agent is trained to solve the task of synthesizing process flowsheets without any heuristics or prior knowledge. The developed RL method models the task as a competitive two-player game that the agent plays against itself during training. The concept is proven to work along an example with a quaternary mixture that is processed using a reactor or distillation units.

New Publication: Automated Process Synthesis Using Reinforcement Learning @ ESCAPE31

New Publication: Automated synthesis of steady-state continuous processes using reinforcement learning

Automated flowsheet synthesis is an important field in computer-aided process engineering. The present work demonstrates how reinforcement learning can be used for automated flowsheet synthesis without any heuristics or prior knowledge of conceptual design. The environment consists of a steady-state flowsheet simulator that contains all physical knowledge. An agent is trained to take discrete actions and sequentially build up flowsheets that solve a given process problem. A novel method named SynGameZero is developed to ensure good exploration schemes in the complex problem. Therein, flowsheet synthesis is modelled as a game of two competing players. The agent plays this game against itself during training and consists of an artificial neural network and a tree search for forward planning. The method is applied successfully to a reaction-distillation process in a quaternary system.