Text

Course syllabus - AI-based Modelling and Optimisation for Industrial Systems, 7.5 credits

Information about the course

  • Course code: FOES032
  • Third-cycle subject: Energy och Environmental Engineering
  • School: School of Business Society and Engineering
  • Responsible department: Department of Sustainable Energy Systems
  • Valid from: Autumn term 2025
  • Established by: Dean of School
  • Decision date: 2025-05-22
  • Last modofied: 2025-05-22
  • Level of education: Third cycle level

Course objective

The purpose of the course is to familiarize students with the basics of Python as a platform to use AI/data-driven methods and optimisation for energy systems. Current practices and future developments will be discussed along with a diverse array of application case studies.

In this course, participants will learn about current state-of-the-art AI-based modelling and optimisation methods, including case studies of relevant applications. Special focus is given to educational applications and modelling, and optimisation for industrial systems.

Course content

  • Introduction to the basics of Python in ML/AI context.
  • Introduction to AI methods for classification/regression, supervised vs unsupervised learning, and regression methods.
  • Artificial neural networks (ANN) for diagnostics, e. g. on ANN-based diagnostics, case studies.
  • Time sequential prediction: Case studies.
  • Fundamentals of Reinforcement Learning with case study.
  • Introduction to optimisation methods, problem classification, figure of merit selection, and analysis of optimisation results - Genetic algorithms and AI-based optimisation.
  • Linear optimisation, production planning, and data-driven control.
  • Transformers.
  • Physics informed Neural Network.
  • Large language Models-1.
  • Large language Models-2.
  • Uncertainty quantification, Monte Carlo dropouts and Deep Ensembles methods.

Intented learning outcomes

After completing the course, the doctoral student should be able to:

  1. Describe basic syntax, computer types and structure in Python.
  2. Develop algorithms in Python code using open-source libraries (numpy, scipy, matplotlib), with the purpose of calculating and presenting results graphically.
  3. Implement the basic principles and applications of AI/data-driven methods.
  4. Describe the basics of Reinforcement Learning, transformers, PINN & LLM models with case studies.
  5. Develop ANN and CNN concepts with case studies.
  6. Describe and develop the basics of uncertainty analysis using Monte Carlo dropouts and Deep Ensembles methods
  7. Identify and implement one of the in the course discussed case studies that best help solving a selected problem.

The intended qualitative targets in relation to the Higher Education Ordinance, appendix 2.

Knowledge and understanding

For the Degree of Doctor, the doctoral student shall demonstrate:

  • A1: broad knowledge and systematic understanding of the research field as well as advanced and up-to-date specialised knowledge in a limited area of this field, and
  • A2: familiarity with research methodology in general and the methods of the specific field of research in particular.

Competence and skills

For the Degree of Doctor, the doctoral student shall demonstrate

  • B1: the capacity for scholarly analysis and synthesis as well as to review and assess new and complex phenomena, issues, and situations autonomously and critically, and
  • B2: the ability to identify and formulate issues with scholarly precision critically, autonomously, and creatively, and to plan and use appropriate methods to undertake research and other qualified tasks within predetermined time frames and to review and evaluate such work.

Judgement and approach

For a Degree of Doctor the doctoral student shall demonstrate

  • C1: intellectual autonomy and disciplinary rectitude as well as the ability to make assessments of research ethics.

Teaching formats

Seminars, lectures and workshops.

Examination

NAR1, active participation at seminars, lectures and workshops, 3 cr, concerning learning outcomes 1, 4, 5 & 6, grade Fail (U) or Pass (G).

INL1, individually written assignment, 4.5 cr, concerning learning outcomes 2, 3 & 7, grade Fail (U) or Pass (G).

For Pass (G) on the course, the participant must pass both NAR1 and INL1.

Grade

Examinations included in the course are assessed according to a two-grade scale: fail or pass.

A person who has not passed the regular examination shall be given the opportunity to retake the test.

Requirements

To participate in the course and the examinations included in the course, the applicant must be admitted to doctoral studies.

Selection criteria

Selection of applicants will be made in accordance with the ranking below.

  1. Doctoral students in Energy Engineering or Environmental Engineering.
  2. Doctoral students at Mälardalen University.
  3. Doctoral students from other universities.

Transitional and other provisions

This course overlaps with FOES005 and FOES019 and cannot be included in the same degree.