Parameter Estimation for Batch Crystallization Processes Using Automatic Differentiation

Weixu Chen, Fangkun Zhang, Tao Liu, Xiongwei Ni

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In this paper, an automatic differentiation-based sequential quadratic programming (AD-SQP) method is proposed for parameter estimation of nonlinear growth models for industrial crystallization processes. Based on the process population balance equation (PBE) related to the size-dependent growth kinetic, the parameter estimation problem is transformed into a constrained optimization problem, such that an automatic differentiation (AD) algorithm is established to efficiently compute the gradients of the model fitting error for parameter estimation. A comparative study between SQP and AD-SQP is made for reference. Simulation results demonstrates that the proposed method is more efficient and accurate than the traditional SQP. Experiments on the seeded batch cooling crystallization of the β form L-glutamic acid (β-LGA) with different temperature cooling profiles are performed to apply the proposed AD-SQP to identify the growth model. The predicted moments of crystal size distribution (CSD) and solution concentration well match with the measured results.

Original languageEnglish
Title of host publicationProceedings of the 37th Chinese Control Conference
EditorsXin Chen, Qianchuan Zhao
PublisherIEEE
Pages1809-1814
Number of pages6
ISBN (Electronic)9789881563958
DOIs
Publication statusPublished - 8 Oct 2018
Event37th Chinese Control Conference 2018 - Wuhan, China
Duration: 25 Jul 201827 Jul 2018

Publication series

NameChinese Control Conference (CCC)
PublisherIEEE
ISSN (Electronic)1934-1768

Conference

Conference37th Chinese Control Conference 2018
Abbreviated titleCCC 2018
CountryChina
CityWuhan
Period25/07/1827/07/18

Keywords

  • Automatic Differentiation
  • Crystallization
  • Parameter Estimation
  • Population Balance Model

ASJC Scopus subject areas

  • Computer Science Applications
  • Control and Systems Engineering
  • Applied Mathematics
  • Modelling and Simulation

Fingerprint Dive into the research topics of 'Parameter Estimation for Batch Crystallization Processes Using Automatic Differentiation'. Together they form a unique fingerprint.

Cite this