Social network you want to login/join with:
Thesis - Self-supervised Learning for Battery Health Estimation f/m/d, Graz
Client: AVL List GmbH
Location: Graz, Austria
Job Category: Other
EU work permit required: Yes
Job Reference:
dac0d1b5455a
Job Views:
3
Posted:
14.03.2025
Expiry Date:
28.04.2025
Job Description:
We are looking for a motivated student to conduct their master thesis in the area of Li-ion battery modelling using state-of-the-art machine learning techniques. This master thesis focuses on developing advanced techniques to estimate battery health and performance in the automotive industry. By leveraging deep neural network architectures for learning the trajectory of degradation with existing test data, the aim is to estimate the state-of-health without having the entire history of the battery’s operation (zero-shot learning). The thesis will contribute to overcoming practical issues for state-of-health estimation in the field and will offer valuable insights into understanding the influencing aging factors.
Responsibilities:
* Literature research: Identify the state-of-art for specific applications and rank the most relevant architectures/techniques.
* Data preparation and pre-processing: Utilize time series analysis and aggregation techniques to create a pipeline for feature engineering during charge cycles, including selection of target variables.
* Data segmentation: Prepare samples of data from existing experimental datasets for training the models.
* Comparison and ablation study: Establish a set of baseline methods (i.e., MLP, RNN, LSTM) for comparison purposes.
* Final model evaluation: Utilize the trained models for final evaluation in both experimental and real-world data.
* Sensitivity analysis: Utilize Explainable-AI methods to pinpoint influencing factors and explain the model’s outputs.
Minimum Requirements:
* BSc in domains similar to Applied Statistics/Mathematics, Computer Science, Data Science, Automotive, or Electrical Engineering.
* Strong background in data analysis, deep learning, and time series prediction.
* Proficiency in programming languages such as Python for implementing data analysis algorithms.
* Familiarity with statistical methods and transformers (LLMs).
* Ability to work independently, conduct experiments, and analyze complex datasets.
* Excellent problem-solving and critical-thinking skills.
* Strong communication skills to present findings and recommendations effectively.
#J-18808-Ljbffr