Logo/Imagem do Projeto
MR2BEST
Project Status
Concluded
Project ID
PTDC/QEQ-EPS/1323/2014
Lead Researcher
Marco Paulo Seabra dos Reis
Funding Agency
FCT - Foundation for Science and Technology
Grant Value
108.252,00 €
Duration
15/06/2016 to 15/12/2019

Process Industries are generating data with a variety of structures that are not properly handled by current process monitoring, control, and optimization approaches. In particular, existing methodologies rely heavily on the availability of data at a single resolution and, most often, at a single acquisition rate. Briefly, resolution is a measure of the degree of information localization for a given quantity over a domain of relevant independent variables (time or space).

For example, some variables represent instantaneous process values (i.e., averages over a very short time period), while others consist of averages over several minutes, hours, or even days; still, others may be measurements representative of a batch produced during the last hour, or of raw materials supplied to the process sometimes over more than 20 days (such as wood chips in the pulp and paper industry). These variables carry information with different temporal localizations—that is, at distinct time resolutions.

However, the common tacit assumption is that all process data are available at the same resolution, usually with high localization in the time domain around the sampling instants (which should, moreover, be equally spaced). Analyzing modern process databases, it is easily verified that this assumption is frequently not met, meaning these databases generally exhibit a multi-resolution data structure. Yet, to the best of our current knowledge, there are no available solutions to handle multi-resolution data even for critical low-level tasks like process monitoring and control, the only exception being a methodology proposed by the PI in 2006 (Reis & Saraiva, 2006c).

As noted in that publication, even so-called multi-scale approaches assume the presence of data at a single resolution, a fact that is often overlooked. A similar situation is found in the higher-level strategic task of process optimization. Therefore, in this project, we aim to develop the conceptual framework and theory for multi-resolution and multi-rate approaches within the context of the following Process Systems Engineering (PSE) activities:

  • Task 1: Soft sensors for stationary (continuous) and non-stationary (batch) dynamic processes.
  • Task 2: Process monitoring for non-stationary (batch) systems.
  • Task 3: Mechanistic process modeling with multi-resolution processing units and the development of suitable optimal estimation tools.
Partner Institutions
UC_resized
University of Coimbra
https://www.uc.pt/
Resultados e Publicações

Journal Articles and Book Chapters

  1. Rendall, R., B. Lu, I. Castillo, S.-T. Chin, L. H. Chiang, M.S. Reis, A Unifying and Integrated Framework for Feature Oriented Analysis of Batch Processes. Industrial & Engineering Chemistry Research. (2017). 56(30), p. 8590-8605.
  2. Reis, M.S., G. Gins, Industrial Process Monitoring in the Big Data/Industry 4.0 Era: From Detection, to Diagnosis, to Prognosis. Processes. 5(3), 35, (2017), p.1-16.
  3. Rato, T.J., M.S. Reis, Improved Fault Diagnosis in Online Process Monitoring of Complex Networked Processes: a Data-Driven Approach. in Computer Aided Chemical Engineering, A. Espuña, M. Graells, and L. Puigjaner, Editors. (2017), Elsevier. p. 1681-1686.
  4. Campos, M., R. Sousa, A.C. Pereira, M.S. Reis, Advanced Predictive Methods for Wine Age Prediction: Part II - A Comparison Study of Multiblock Regression Approaches. Talanta. 171 (2017), p. 132-142.
  5. Rendall, R., A.C. Pereira, M.S. Reis, Advanced Predictive Methods for Wine Age Prediction: Part I - A Comparison Study of Single-Block Regression Approaches based on Variable Selection, Penalized Regression, Latent Variables and Tree-based Ensemble Methods. Talanta. 171 (2017), p. 341-350. 
  6. Rato, T.J., M.S. Reis, Multiresolution Soft Sensors: A New Class of Model Structures for Handling Multiresolution Data. Industrial & Engineering Chemistry Research. 56(13) (2017), p. 3640-3654.
  7. Rato, T.J., M.S. Reis, Markovian and Non-Markovian Sensitivity Enhancing Transformations for Process Monitoring. Chemical Engineering Science. 163 (2017), p. 223-233.
  8. Rendall, R. and M.S. Reis, Which regression method to use? Making informed decisions in “data-rich/knowledge poor” scenarios – The Predictive Analytics Comparison framework (PAC). Chemometrics and Intelligent Laboratory Systems. 181 (2018), p. 52-63.
  9. Rato, T.J. and M.S. Reis, Optimal selection of time resolution for batch data analysis. Part I: Predictive modeling. AIChE Journal. 64 (2018), p. 3923-3933.
  10. Reis, M.S., R.S. Kenett, Assessing the Value of Information of Data-Centric Activities in the Chemical Processing Industry 4.0. AIChE Journal. 64 (2018), p. 3868-3881.
  11. Rato, T.J. and M.S. Reis, Building Optimal Multiresolution Soft Sensors for Continuous Processes. Industrial & Engineering Chemistry Research. 57 (2018), p. 9750-9765.
  12. Rato, T.J., R. Rendall, V. Gomes, P.M. Saraiva, and M.S. Reis, A Systematic Methodology for Comparing Batch Process Monitoring Methods: Part II—Assessing Detection Speed. Industrial & Engineering Chemistry Research. 57(15) (2018), p. 5338-5350.
  13. Campos, M.P., R. Sousa, M.S. Reis, Establishing the Optimal Blocks’ Order in SO-PLS: Stepwise SO-PLS and Alternative Formulations. Journal of Chemometrics. 32 (2018), p. e3032.
  14. Geert, G., J. Van Impe, M.S. Reis, Finding the optimal time resolution for batch-end quality prediction: MRQP – a framework for Multi-Resolution Quality Prediction. Chemometrics and Intelligent Laboratory Systems. 172 (2018), p. 150-158.
  15. Santos, C.P., T.J. Rato, and M.S. Reis, Design of Experiments: A Comparison Study from the Non-Expert User’s Perspective. Journal of Chemometrics. (2018), p. e3087.
  16. Rato, T.J. and M.S. Reis, Optimal fusion of industrial data streams with different granularities. Computers & Chemical Engineering. 130 (2019), p. 106564.
  17. Rato, T.J. and M.S. Reis, SS-DAC: A systematic framework for selecting the best modeling approach and pre-processing for spectroscopic data. Computers & Chemical Engineering. 128 (2019), p. 437-449.
  18. Rato, T.J. and M.S. Reis, Multiresolution interval partial least squares: A framework for waveband selection and resolution optimization. Chemometrics and Intelligent Laboratory Systems. 186 (2019), p. 41-54.
  19. Reis, M.S., Multiscale and Multi-granularity Process Analytics: A Review. Processes. 7 (2) (2019), p. 61
  20. Rendall, R., L.H. Chiang, M.S. Reis, Data-driven Methods for Batch Data Analysis – A Critical Overview and Mapping on the Complexity Scale. Computers & Chemical Engineering. 124 (2019), p. 1-13.
  21. Reis, M.S., G. Gins, and T.J. Rato, Incorporation of process-specific structure in statistical process monitoring: A review. Journal of Quality Technology. (2019), p. 1-15.
  22. Reis, M.S. and T.J. Rato, An Advanced Data-Centric Multi-Granularity Platform for Industrial Data Analysis, in Computer Aided Chemical Engineering, A.A. Kiss, et al., Editors. 2019, Elsevier. p. 1225-1230.

 

International Conference Communications

  1. Reis, M.S., Advances in Batch Data Analysis. Seminário apresentado na conferência “AIChE Spring Meeting”, realizada em San Antonio (EUA), entre 26 a 30 de março de 2017.
  2. Reis, M.S., Predictive Modeling with High-Dimensional Industrial Data, Apresentação oral em formato de seminário apresentada na conferência “JDS 2017 – 49èmes Journées de Statistique”, realizada em Avignon (França), entre 20 de maio e 2 de junho de 2017. 
  3. Reis, M.S., On the importance of residual analysis in the implementation of PCA and PLS, Comunicação apresentada no formato de poster no “JMP Discovery Summit”, realizado em Praga (República Checa), entre 21 e 23 de março de 2017.
  4. Gins, G.; J. Van Impe, M.S. Reis, M.R.Q.P.: Prediction of Final Batch Quality Using a Multi-Resolution Framework, Comunicação oral apresentada no congresso “2016 AICHE Annual Meeting”, realizado em San Franciso (CA, EUA), entre 13 e 18 de novembro de 2016.
  5. Rendall, R., B. Lu, I. Castillo, S.-T.-Chin, L.H. Chiang, M.S. Reis, Parsimonious Modeling Approaches for Batch Process Analysis, Comunicação oral apresentada no congresso “2016 AICHE Annual Meeting”, realizado em San Franciso (CA, EUA), entre 13 e 18 de novembro de 2016.
  6. Reis, M.S., Structured Approaches for High-Dimensional Predictive Modeling, Comunicação oral apresentada na conferência “SIS2017 - Statistics and Data Science: New Challenges, New Generations”, realizada em Florença (Itália), entre 28 e 30 de junho de 2017.
  7. Rendall, R., B. Lu, I. Castillo, S.-T.-Chin, L.H. Chiang, M.S. Reis, Profile-driven Features for Offline Quality Prediction in Batch Processes. Comunicação apresentada no formato de poster no congresso “ESCAPE-27, European Symposyum on Computer Aided Process Engineering”, realizado em Barcelona (Espanha), entre 1 e 5 de outubro de 2017.
  8. Reis, M.S., R.S. Kenett, On the Use of Simulators for Teaching Statistical Methods. Comunicação oral apresentada no congresso “ENBIS16 – 16th Annual ENBIS Conference”, realizado em Sheffield (UK), entre 11 e 15 de setembro de 2016.
  9. Rato, T.J., M.S. Reis, Markovian and Non-Markovian Sensitivity Enhancing Transformations for Process Monitoring. Comunicação oral apresentada no congresso “ENBIS16 – 16th Annual ENBIS Conference”, realizado em Sheffield (UK), entre 11 e 15 de setembro de 2016.
  10. Rato, T.J. and M.S. Reis. Improved Fault Diagnosis in Online Process Monitoring of Complex Networked Processes: a Data-Driven Approach. Comunicação oral apresentada no congresso “27th European Symposium on Computer Aided Process Engineering”, realizado em Barcelona (Espanha), entre 1 e 5 de Outubro de 2017.
  11. Rato, T.J. and M.S. Reis. A Multiresolution Framework for Building Industrial Soft Sensors. Comunicação oral apresentada no congresso “ENBIS18 – 18th Annual ENBIS Conference”, realizado em Nancy (França), entre 2 e 6 de setembro de 2018.
  12. Reis. M.S., A Systematic Framework for Assessing the Quality of Information in Data-Driven Applications for the Industry 4.0. Comunicação oral apresentada no congresso “ADCHEM 2018, 10th IFAC Sumposium on Advanced Control of Chemical Processes ”, realizado em Shenyang (China), entre 25 e 27 de julho de 2018. (Inclui artigo publicado nos proceedings do congresso).
  13. Reis. M.S., T.J. Rato, Multiresolution Analytics for Large Scale Industrial Processes. Comunicação oral apresentada no congresso “ADCHEM 2018, 10th IFAC Sumposium on Advanced Control of Chemical Processes ”, realizado em Shenyang (China), entre 25 e 27 de julho de 2018. (Inclui artigo publicado nos proceedings do congresso).
  14. Reis, M.S., Incorporating Systems Structure in Data-Driven High-Dimensional Predictive Modeling. Comunicação oral apresentada no congresso “ESCAPE-28, European Symposyum on Computer Aided Process Engineering”, realizado em Graz (Áustria), entre 10 e 13 de junho de 2018.
  15. Reis, M.S., Exploring the Latent Variable Space of a Multiresponse DOE to Optimize Solid Phase Microextraction (SPME): Case study - Quantification of Volatile Fatty Acids in Wines. Comunicação oral apresentada no congresso “ENBIS Spring Meeting on Design of Experiments for Quality of Products and Sustainability in Agri-Food Systems”, realizado em Florença (Itália), entre 4 e 6 de junho de 2018.
  16. Reis, M.S., Process Analytics for Quality Improvement. Palestra realizada por convite no congresso internacional da European Organization for Quality (EOQ) 2019, realizado em Lisboa, entre 23 e 24 de outubro de 2019.
  17. Reis, M.S., Industrial Data Science for Quality Improvement. Palestra realizada por convite na 17th Workshop on Quality Improvement Methods (Dortmund, Alemanha) entre 14 e 15 de junho de 2019.
  18. Reis, M.S., Modern Approaches to Industrial Process Monitoring. Comunicação oral apresentada no congresso “ENBIS19 – 19th Annual ENBIS Conference”, realizado em Budapeste (Hungria), entre 2 e 4 de setembro de 2019.
  19. Reis, M.S., T.J. Rato, An Advanced Data-Centric Multi-Granularity Platform for Industrial Data Analysis. Comunicação oral apresentada no congresso “ESCAPE-29, European Symposyum on Computer Aided Process Engineering”, realizado em Eindhoven (Holanda), entre 16 e 19 de junho de 2019.
  20. Reis, M.S., Structured data-driven approaches for process monitoring and predictive analytics. Palestra realizada por convite na Hybrid Modeling Summer School, que teve lugar na Universidade Nova de Lisboa, no dia 27 de setembro de 2019.

 

Software

  • Multiresolution Soft Sensors (MR-SS)
  • Multiresolution Kalman Filter (MR-KF)
  • Multiresolution Empirical Models for Continuous Processes (MR-EMC)
  • Multiresolution Models for Batch Processes (MR-PLS)
  • Multiresolution Batch Process Monitoring (MR-BPM)
  • Multiresolution interval Partial Least Squares (MR-iPLS)
  • Soft Sensor Development, Assessment and Comparison (SS-DAC)
  • Integrated package for Multiresolution Modelling

 

PT2020