ETI 06-2023 Master thesis State-of-Energy estimation using Machine Learning considering dynamic current rates and temperatures

State estimation of Energy Storage Systems (ESS) is really important for its safe and optimal operation. State estimation could be of various kinds of which State-of-Charge (SOC) and State-of-Health (SOH) is what is commonly researched. For energy management of ESS, it is important to estimate the amount of energy that could be taken out of the ESS. This is termed as State-of-Energy (SOE) of ESS. The energy capacity varies according to the operation conditions, as can be seen in the following graph.

Under higher current rate and/or higher temperature the ESS cannot deliver the defined nominal energy capacity. Thus, under dynamic current rates and temperatures the SOE has to be actively updated, based on which an optimal energy management strategy could be defined.

SOE estimation works based on ESS models. But unfortunately, ESS modelling is not straight forward and SOE estimation under external factors such as balancing, uneven temperature distribution in ESS, and others could make it a complex task overall. One of the ways to solve this problem is by utilizing Machine Learning (ML) techniques to build a black box model of the system based on real operation data. This is the aim of the advertised master thesis. During the master thesis various ML techniques could be applied on 1-year data (250msec resolution) and SOE estimation models could be developed and compared on real operation of the system.

Persönliche Qualifikation

Courses of Study: Electrical Engineering, Mechanical Engineering, Mathematics, Process Engineering or Informatics

Focus: Theory, Literature, Simulation and Programming

Skills recommended: MATLAB, Simulink, basic understanding of battery, Cell Modelling, machine learning

Organisationseinheit

Contract duration

6 months

Contact person in line-management

For further information, please contact Mister Palaniswamy, email: lakshimi.palaniswamy@kit.edu.