De elektronische studiegids voor het academiejaar 2025 - 2026 is onder voorbehoud.





New strategies for process control (4474)
  
Coordinating lecturer :Prof. dr. ir. Mumin enis LEBLEBICI 
  
Co-lecturer :Prof. dr. ir. Jozefien DE KEYZER 

Language of instruction : English
Credits: 4,0
  
Period: semester 1 (4sp)
  
2nd Chance Exam1: Yes
  
Final grade2: Numerical
 
Exam contract: not possible
 
Sequentiality
 
   Advising sequentiality bound on the level of programme components
 
 
Group 1
 
  Following programme components are advised to also be included in your study programme up till now.
    Chemical design and safety (3313) 4.0 stptn
    ICT tools for Chemical Engineers (4033) 3.0 stptn
    Process control (2624) 3.0 stptn
    Reactor engineering (3301) 3.0 stptn
    Separation processes (4335) 3.0 stptn
 
Or group 2
 
  Following programme components are advised to also be included in your study programme up till now.
    Chemical design and safety (3313) 4.0 stptn
    Coding & Scripting transition (4293) 3.0 stptn
    ICT tools for Chemical Engineers (4033) 3.0 stptn
    Process control (2624) 3.0 stptn
    Reactor engineering (3301) 3.0 stptn
    Separation processes (4335) 3.0 stptn
 

Content

DATA DRIVEN MODELLING AND CONTROL

1.INTRODUCTION

  • Course introduction: teaching group, course material, suggested videos and books, timeline, evaluation method
  • AI introduction: what is AI and its advantages, the categories of AI and applications, optimization in chemical engineering, AI methods applied to modeling, AI methods applied to control, AI methods applied to optimizations
  • Python introduction: Python documents and learning materials, introduction to environments and IDEs, Python installation, introduction to python libraries used during the course

2.DATA PROCESSING

    • Scaling techniques: scaling data to a range, scaling sparse data, scaling data with outliers, normalization
  • Data cleaning from industrial processes: how a database is done, data flow on a plant, case specific measurement issues and how to overcome them, data filtering

3.BLACK BOX MODELLING

  • Artificial Neural network: applications in chemical engineering, activation functions, multilayer perceptron, backpropagation algorithm, overfitting and early stopping, cross-validation
  • Multivariate rational function: approximation of a function, ideal distribution for approximation, distribution in chemical engineering case, the structure of an multivariate rational function, difference between polynomial and multivariate rational function
  • Heuristic optimization: particle swarm optimization and application in chemical engineering, differential evolution and application in chemical engineering
  • Gradient-based optimization: differences with an heuristic optimization, gradient descent, stochastic gradient descent, learning rate, linear programming

4.HYBRID MODELLING

  • Build and application of hybrid model in chemical case: different stucture of an hybrid model, the advantages to use an hybrid model rather a pure black-box or white-box, hybrid model application in chemical engineering

5.CONTROL

  • Model-predictive control

6.AI IN CHEMICAL CASES

  • Reaction case
  • Separation case


Pinch technology: introduction & solving problems using Excel


Organisational and teaching methods
Organisational methods  
Application Lecture ✔  
Small group session ✔  
Teaching methods  
Exercises ✔  
Group work ✔  
Homework ✔  


Evaluation

Period 1    Credits 4,00

Evaluation method
Written evaluaton during teaching periode90 %
Transfer of partial marks within the academic year✔
Conditions transfer of partial marks within the academic yearOne Test using Excel (pinch 25%) One report(group 50%) 2 Homeworks (individual 15%) The grade for the homework tasks will be retained. The test on Pinch technology (25% of Final Grade - FG) can be retaken if the grade is < 12/20. If the grade on this test is > 12/20, this grade will also be retained. Almost every week there are very small homeworks. There will be 2 larger individual homeworks during the semester. Homework grade retained to 2nd exam. At the end of the semester the students will deliver reports. (Group work /50% FG). Report grade retained for second chance. This report will also be defended via a presentation (see evaluation in exam eriod 10% FG).
Closed-book ✔
Homework✔
Open questions✔
Report✔
Oral exam10 %
Transfer of partial marks within the academic year✔
Presentation✔
Additional information

Almost every week there are very small homeworks. There will be 2 larger individual homeworks during the semester. (Average of homeworks - larger : 15% of final grade-FG) At the end of the semester the students will deliver reports. (Group work /50% FG) This report will also be defended via a presentation (10% FG).

There will also be a test for the part on Pinch technology (scored practice session, closed book with open formulas 25% FG) 

Second examination period

Evaluation second examination opportunity different from first examination opprt
No
Explanation (English)The report (50% Final Grade-FG) can be corrected and send as second chance. The test (25% FG) can be repeated. The other grades remain the same, no second chance for that part.
 

Recommended course material
 

1. Links:

2. Books:

  • “Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow” by Aurelien Geron
  • “Python for Data Analysis: Data Wrangling with Pandas, NumPy, and IPython” by Wes McKinney
  • “Hybrid Modeling in Process Industries” by Jarka Glassey
  • “Artificial intelligence in chemical engineering” by Thomas E. Quantrille

3. Papers:

  • Venkatasubramanian, V. The Promise of Artificial Intelligence in Chemical Engineering: Is It Here, Finally? (2018) doi:10.1002/aic.16489.
  • von Stosch, M., Oliveira, R., Peres, J. & Feyo de Azevedo, S. Hybrid semi-parametric modeling in process systems engineering: Past, present and future. Comput. Chem. Eng. 60, 86–101 (2014).
  • Chiang, L., Lu, B. & Castillo, I. Big Data Analytics in Chemical Engineering. Annu. Rev. Chem. Biomol. Eng. 8, 63–85 (2017).

4. Software: 

  • Excel



Chemic al process: design and integration, Robin Smith, Wiley, 9781119094418

Documents on electronic learning platform (Toledo)
 

Remarks
 

Orientation in the curriculum: 

This course is part of the learning domain Proces design and engineering in the optie duurzame procestechnologie en kunststoffen.


Learning outcomes
Master of Chemical Engineering Technology
  •  EC 

    • EC2 - The holder of the degree masters a comprehensive set of chemical techniques and technologies and is able to creatively conceptualise, plan and execute these as an integrated part of a methodologically and systematically ordered series of actions within a multidisciplinary project with a significant research and/or innovation component.

     
  •  DC 

    • DC1 - The student has knowledge of the basic concepts, structures and coherence.

      
  •  BC 
    		• Student is able to understand the theory behind hybrid modelling, its techniques and tools.
     
  •  DC 

    • DC3 - The student can recognize problems, plan activities and perform accordingly.

      
  •  BC 
    		• Student is able to understand when the AI is a viable tool.

    Student is able to select the adequate tool for modelling and control.
     
  •  DC 

    • DC4 - The student can gather, measure or obtain information and refer to it correctly.

      
  •  BC 
    		• Student is able to locate and use the adequate algorithms/libraries for modelling and control.
     
  •  DC 

    • DC7 - The student can use selected methods and tools to implement solutions and designs.


      
  •  BC 
    		• Student is able to implement the selected AI techniques for chemical process modelling and control problems.
  •  EC 

    • EC5 - The holder of the degreehas advanced or specialist knowledge of and insight in most unit operations in the (bio)chemical industry and can integrate this knowledge and insight to creatively conceptualise and autonomously control and simulate chemical processes and to develop process optimisations within a multidisciplinary design context and with attention to topical technological developments and innovations.

     
  •  DC 

    • DC1 - The student has knowledge of the basic concepts, structures and coherence.

      
  •  BC 
    		• The student knows the rules of pinch technology and can apply them.
     
  •  DC 

    • DC2 - The student has insight in the basic concepts and methods.

      
  •  BC 
    		• Student is able to criticize the response of the AI and can troubleshoot it.

    The student shows insight in the application of pinch technology rules.
     
  •  DC 

    • DC3 - The student can recognize problems, plan activities and perform accordingly.

      
  •  BC 
    		• Student is able to recognize adequate scenarios/problems to implement AI techniques.
     
  •  DC 

    • DC5 - The student can analyze problems, logically structure and interpret them.

      
  •  BC 
    		• The student analyses the given process in order to design a heat exchanger network and analyses the obtained results.
     
  •  DC 

    • DC6 - The student can select methods and make calculated choices to solve problems or design solutions.


      
  •  BC 
    		• The student can design a heat exchanger network for a given process.
  •  EC 

    • EC6 - The holder of the degree has advanced or specialist knowledge of, insight in and proficiency within a self-selected domain of specialisation of (bio)chemical process technology, materials, food and/or packaging.

     
  •  DC 

    • DC1 - The student has knowledge of the basic concepts, structures and coherence.

      
  •  BC 
    		• The data driven modelling tools, heuristics, code syntax is learned.
     
  •  DC 

    • DC2 - The student has insight in the basic concepts and methods.

      
  •  BC 
    		• The student can estimate which algorithm and solvers to use and can code accordingly with and without AI support.
     
  •  DC 

    • DC3 - The student can recognize problems, plan activities and perform accordingly.

      
  •  BC 
    		• The student can estimate which algorithm and solvers to use, which heuristic optimizer to apply and can code accordingly with and without AI support.
 
  EC = learning outcomes      DC = partial outcomes      BC = evaluation criteria  
Offered inTolerance3
Exchange Programme Engineering Technology J
Master of Chemical Engineering Technology optie duurzame procestechnologie J



1   Education, Examination and Legal Position Regulations art.12.2, section 2.
2   Education, Examination and Legal Position Regulations art.15.1, section 3.
3   Education, Examination and Legal Position Regulations art.16.9, section 2.