Smart Battery II: Artificial Intelligence in Battery State Estimation (2024)

Organizer:     Postdoc Xin Sui xin@energy.aau.dk & Prof. Remus Teodorescu ret@energy.aau.dk  

ECTS: 2.0

Date/Time: 27-28 November 2024

Deadline: 06 November 2024

Place: AAU Energy, Pontoppidanstraede 111, Aalborg, Denmark

Max no. of participants: 30

Description:  Lithium-ion batteries have a wide range of applications, and their safe and reliable operation is essential. However, due to the complex electrochemical reaction of the battery, the battery performance parameters show strong nonlinearity with aging. Therefore, as the main technologies in BMS, battery state estimation and lifetime prediction remain challenges. Artificial Intelligence (AI) technologies possess immense potential in inferring battery state and can extract aging information (i.e., health indicators) from measurements and relate them to battery performance parameters, avoiding a complex battery modeling process. Therefore, this course aims to introduce the application of AI in Smart Battery state estimation.

This two-day course introduces AI methods for estimating/predicting batteries’ state of charge (SOC), state of health (SOH), state of temperature (SOT), and remaining useful life (RUL). Key aspects include laboratory data preparation, data preprocessing, AI model training and selection.

In addition to the classic algorithms of AI, e.g., support vector regression, Gaussian process regression, neural networks, transfer learning, and multitask learning, feature extraction and selection methods will be included in the discussion.

In terms of training, two modes will be introduced (depending on the accuracy, robustness, and computation complexity of the selected AI algorithm), i.e., with feature extraction and without feature extraction. According to multiple case studies, the strengths and drawbacks of different AI algorithms will be compared.

Exemplifications of some of the discussed topics will be made through exercises in Python and MATLAB.

Day 1: Introduction to AI and battery state estimation

• Lecture 1: Introduction to Smart Battery: how AI makes battery smart  - Remus Teodorescu
• Lecture 2: Data analysis and AI in general part 1 - Nicolai André Weinreich
• Lecture 3: Data analysis and AI in general part 2 - Nicolai André Weinreich
  
• Exercise 1: A demo of classification/regression algorithms - Nicolai André Weinreich
• Lecture 4: Lithium-ion batteries basis - Xin Sui
  
• Lecture 5: Introduction to SOX (SOC, SOT, SOH, RUL) - Xin Sui
• Lecture 6: SOH estimation using machine learning - Xin Sui
  
• Exercise 2: Training an NN for SOC/SOH estimation - Xin Sui

 Day 2: Applied AI in battery modeling and balancing

• Lecture 7: AI for battery modeling and optimization Part 1 - Roberta Di Fonso
• Lecture 8: AI for battery modeling and optimization Part 2 - Roberta Di Fonso
• Exercise 3: Battery modeling in Simulink - Roberta Di Fonso
  
• Lecture 9:  State-of-the-art AI algorithms in battery lifetime prediction - Xin Sui
• Lecture 10: Physics-informed Machine Learning  - Yusheng Zheng
  
• Lecture 11: Physics-informed Machine Learning - Yusheng Zheng
• Wrap up of the course and lab visit

Prerequisites: Fundamental understanding of characteristics of Li-ion batteries, and familiar with programming using MATLAB/Python. Note: the course language is English.

Form of evaluation: Students are expected to solve a few exercises and deliver an individual report with solutions and comments

Price:  6000 DKK for PhD students outside of Denmark and 8000 DKK for the Industry excl. VAT.
A link for online payment by credit card for externals participants will be annonced after deadline for registration

Find more information about the course (more details and course litterature)

Important information concerning PhD courses: We have over some time experienced problems with no-show for both project and general courses. It has now reached a point where we are forced to take action. Therefore, the Doctoral School has decided to introduce a no-show fee of DKK 3000 for each course where the student does not show up. Cancellations are accepted no later than 2 weeks before start of the course. Registered illness is of course an acceptable reason for not showing up on those days. Furthermore, all courses open for registration approximately four months before start. This can hopefully also provide new students a chance to register for courses during the year. We look forward to your registrations.