DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: A novel implicit hybrid machine learning model and its application for reinforcement learning

Journal Article · · Computers and Chemical Engineering
 [1];  [2];  [2]
  1. University of Utah, Salt Lake City, UT (United States); OSTI
  2. University of Utah, Salt Lake City, UT (United States)

A novel methodology to develop implicit hybrid models is presented. PyTorch is used to integrate physics-based equations with machine learning models. Automatic differentiation of the hybrid model is leveraged to solve the implicit equations. Iterative solving enables gradient based updates to the machine learning model. The novel methodology is compared to an explicit hybrid approach on a continuously stirred tank reactor (CSTR). The novel method results in a lower modelling error. Both hybrid models effectively train with noisy data. To test the implicit hybrid model, it is employed as a reinforcement learning (RL) training model. The RL algorithm trained on the hybrid model outperforms real time optimization of the CSTR and performs nearly as well as RL trained directly on the CSTR and a traditional gradient based approach. Training RL directly on the CSTR requires over 60,000 system interactions compared to 6000 historical data points for hybrid model development.

Research Organization:
University of Utah, Salt Lake City, UT (United States)
Sponsoring Organization:
USDOE Office of Energy Efficiency and Renewable Energy (EERE)
Grant/Contract Number:
EE0007712
OSTI ID:
1977010
Journal Information:
Computers and Chemical Engineering, Journal Name: Computers and Chemical Engineering Journal Issue: C Vol. 155; ISSN 0098-1354
Publisher:
ElsevierCopyright Statement
Country of Publication:
United States
Language:
English

References (25)

A framework of hybrid model development with identification of plant‐model mismatch journal August 2020
A hybrid neural network-first principles approach to process modeling journal October 1992
Gray-Box Modeling for the Optimization of Chemical Processes journal November 2018
How to be a gray box: dynamic semi-physical modeling journal November 2001
Applications of hybrid models in chemical, petroleum, and energy systems: A systematic review journal October 2018
A generic hybrid model development for process analysis of industrial fixed-bed catalytic reactors journal January 2017
Optimization and control of a thin film growth process: A hybrid first principles/artificial neural network based multiscale modelling approach journal November 2018
Advances and opportunities in machine learning for process data analytics journal July 2019
Reinforcement learning for batch bioprocess optimization journal February 2020
Deep hybrid modeling of chemical process: Application to hydraulic fracturing journal March 2020
Hybrid Modeling in the Era of Smart Manufacturing journal September 2020
A review On reinforcement learning: Introduction and applications in industrial process control journal August 2020
Real-time optimization using reinforcement learning journal December 2020
Gray-Box system modeling using symbolic regression and nonlinear model predictive control of a semibatch polymerization journal March 2021
Real-time optimization of an industrial steam-methane reformer under distributed sensing journal September 2016
Gray-box modeling and control of polymer molecular weight distribution using orthogonal polynomial neural networks journal October 2012
Gray-box modeling for prediction and control of molten steel temperature in tundish journal April 2014
Process structure-based recurrent neural network modeling for model predictive control of nonlinear processes journal May 2020
Physics-informed reinforcement learning optimization of nuclear assembly design journal February 2021
Construction of a Valid Domain for a Hybrid Model and Its Application to Dynamic Optimization with Controlled Exploration journal August 2020
Partial Observations and Conservation Laws: Gray-Box Modeling in Biotechnology and Optogenetics journal November 2019
Tuning the molecular weight distribution from atom transfer radical polymerization using deep reinforcement learning journal January 2018
Machine learning vs. hybrid machine learning model for optimal operation of a chiller journal September 2018
Development of a hybrid model for a partially known intracellular signaling pathway through correction term estimation and neural network modeling journal December 2020
Comparing Reinforcement Learning Methods for Real-Time Optimization of a Chemical Process journal November 2020