Search Menu
DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information
Search Scholarly Publications

Title: Reinforcement learning for real-time process control in high-temperature superconductor manufacturing

Journal Article · · International Journal of Advanced Manufacturing Technology
 [1];  [1];  [1]; ORCiD logo [2];  [3];  [3];  [3];  [3]
  1. Univ. of Houston, TX (United States)
  2. Univ. of Houston, TX (United States); Texas Center for Superconductivity, Houston, TX (United States); Princeton Plasma Physics Laboratory (PPPL), Princeton, NJ (United States)
  3. Univ. of Houston, TX (United States); Texas Center for Superconductivity, Houston, TX (United States)
+ Show Author Affiliations

With high efficiency and low energy loss, high-temperature superconductors (HTS) have demonstrated their profound applications in various fields, such as medical imaging, transportation, accelerators, microwave devices, and power systems. The high-field applications of HTS tapes have raised the demand for producing cost-effective tapes with long lengths in superconductor manufacturing. However, achieving the uniform and enhanced performance of a long HTS tape is challenging due to the unstable growth conditions in the manufacturing process. Although it is confirmed that the process parameters during the advanced metal organic chemical vapor deposition (A-MOCVD) process influence the uniformity of the produced HTS tapes, the high-dimensional process parameter signals and their complicated interactions make it difficult to develop an effective control policy. In this paper, we propose a local measure for the uniformity of HTS tapes to provide instant feedback for our control policy. Then, we model the manufacturing of HTS tapes as a Markov decision process (MDP) with continuous state and action spaces to assess the instant reward in real time in our feedback control model. As our MDP involves continuous and high-dimensional state and action spaces, a neural fitted Q-iteration (NFQ) algorithm is adopted to solve the MDP with artificial neural network (ANN) function approximation. The collinearity of process parameters can restrict our capability of adjusting the process parameters, which is addressed by the principal component analysis (PCA) in our method. The control policy adjusts the PCA of process parameters using the NFQ algorithm. In conclusion, based on our case studies on real A-MOCVD dataset, the obtained control policy increases the average uniformity of tapes by 5.6% and performs especially well on sample HTS tapes with a low uniformity.

Research Organization:
Princeton Plasma Physics Laboratory (PPPL), Princeton, NJ (United States)
Sponsoring Organization:
USDOE; National Science Foundation (NSF)
Grant/Contract Number:
AC02-09CH11466
OSTI ID:
2282126
Journal Information:
International Journal of Advanced Manufacturing Technology, Journal Name: International Journal of Advanced Manufacturing Technology Journal Issue: 5-6 Vol. 129; ISSN 0268-3768
Publisher:
SpringerCopyright Statement
Country of Publication:
United States
Language:
English

References (24)

Review of research and measurement for application properties of HTS tapes journal July 2014
Critical Current Control in YBa 2 Cu 3 O 7-δ Films Using Artificial Pinning Centers journal January 2005
Effect of deposition temperature on the superconducting properties of Y-Ba-Cu-O films prepared by CVD journal December 1989
Applications of high-temperature superconductors in power technology journal October 2003
Control of the critical current density inYBa2Cu3O7−δfilms by means of intergrain and intragrain correlated defects journal September 1999
Application of high temperature superconductors for fusion journal October 2011
The critical current of superconductors: an historical review journal September 2001
Superconductivity at 93 K in a new mixed-phase Y-Ba-Cu-O compound system at ambient pressure journal March 1987
Neural Fitted Q Iteration – First Experiences with a Data Efficient Neural Reinforcement Learning Method book January 2005
Over 15 MA/cm2 of critical current density in 4.8 µm thick, Zr-doped (Gd,Y)Ba2Cu3Ox superconductor at 30 K, 3T journal May 2018
Growth of High-Performance 4-5μm Thick Film REBCO Tapes Doped With Hafnium Using Advanced MOCVD journal August 2021
Length Uniformity of the Angular Dependences of Ic and n of Commercial REBCO Tapes with Artificial Pinning at 77 K journal January 2019
Thickness dependence of structure and superconductivity of the SmBa2Cu3O7 film by laser CVD journal January 2017
Differences between two definitions of the critical current of HTS coils journal October 2013
High Performance 2G-HTS Wire Using a Novel MOCVD System journal June 2015
Reinforcement learning with Gaussian processes for condition-based maintenance journal August 2021
Fabrication of Coated Conductors by Multiple-Stage CVD journal June 2005
Role of grain size in superconducting boron-doped nanocrystalline diamond thin films grown by CVD journal December 2011
A New High- T c Oxide Superconductor without a Rare Earth Element journal February 1988
Impact of critical current fluctuations on the performance of a coated conductor tape journal October 2019
Novel technologies and configurations of superconducting magnets for MRI journal July 2013
Current non-uniformity in multistrand superconducting cables. Experimental studies and its influence on stability of superconducting magnets journal March 2000
Possible highT c superconductivity in the Ba?La?Cu?O system journal June 1986
Critical current density above 15 MA cm −2 at 30 K, 3 T in 2.2 μ m thick heavily-doped (Gd,Y)Ba 2 Cu 3 O x superconductor tapes journal May 2015

Similar Records

Related Subjects

75 CONDENSED MATTER PHYSICS
SUPERCONDUCTIVITY AND SUPERFLUIDITY
High-temperature superconductor
Reinforcement learning
Function approximation
Artificial neural network
Process control