Thesis Worker - Neural networks for on-line modeling of BEV electrical parameters and losses i Göteborg

Thesis Worker at Volvo Cars

Welcome to explore the world of Volvo Cars by writing your thesis with us! As a thesis worker in our organization you are supported by a supervisor who follows you during your project. All thesis projects are arranged in business critical areas and therefore you will be able to contribute to our company purpose – providing freedom to move in a safe, sustainable and personal way – from day one! 

Background 
The Electric Drive department 97700 at Volvo Cars is proud of being responsible for designing and delivering system solutions and functionality for the inverters, electric machine, and transmission, in our electrified vehicles. 
 

Our vision is to become “Leading in electrification” and we are aiming to contribute to world class energy efficiency, E-machine functionality, sustainability, customer satisfaction and best performance and drivability, to all our future products.

Scope
An accurate assessment of an E-machine’s electrical characteristics and losses is critical in predicting system performance and ensuring a safe and reliable system operation. 
 

The methods that can be used to estimate the losses in an E-machine are:
- Analytical method
Inductances are modelled as functions of measured currents. For a given torque and a given speed, the loss minimizing current sets are constructed into a lookup table (LUT). 
- Temperature Mapping and Correlation with Thermal model
A 1D-thermal model is constructed with calculated losses assigned for various components of the EM.
- Artificial Neural Networks (ANN)
A well-defined and properly trained ANN can accurately predict the EM’s losses. Using Design of Experiment (DoE) methods, we can properly define the main ANN parameters that a minimum prediction error could be reached.

Activities we foresee that should be done during this thesis work are: 
- Literature study on the EM losses and respective methods for predicting them. Literature study on different ANN models used in prediction tasks.
- Create an ANN to predict the electrical characteristics and losses in EMs, based on measured values of currents, voltages, and temperatures. 
- Integrate the losses, predicted by ANN, to the 1D-Thermal model to estimate the respective temperatures. Compare the results with those estimated by losses, which have been calculated analytically.
- Develop an ANN to model a LUT. Apply DoE methods for optimum definition of the parameters and their values in the LUT.
- Compare the LUT model created with ANN and DoE, to the LUT created by inductances based on both modelled and measured data.
- Production of final report.

Your contribution is our future 
You are studying either M.Sc. Electrical engineering, or M.Sc. Automation engineering, or similar. Understanding in Neural Networks is highly valuable. Good knowledge of MATLAB and Python is necessary. 
 

Duration 
• The duration of this thesis work is approx. 4-5 months. 
• Starting date is in January 2023, exact date to be discussed. 
• 30 ECTS 
• This thesis work is to be conducted by 2 students working in pair. 
• Office, computer and software will be organized by Volvo Cars, in close co-operation with your university.
 

Be part of the change – apply today!
Selection will be ongoing during the application period, so do not hesitate to send in your application. Attach your resume and cover letter and copies of grades. You can provide a reference if you like. 
 

Please note that due to GDPR, applications by email will not be accepted. 
 

We want your application at the latest by 2022-11-18 

If you have any additional questions regarding the position, you are welcome to contact us. 
Supervisor Kristian Holmberg, kristian.holmberg@volvocars.com (+46 729 70 72 04)
Hiring manager Anders Thulin, anders.thulin@volvocars.com (+46 723 71 60 43)

We are Volvo Cars

Volvo Cars is a company on a mission to bring traditional car manufacturing into a connected, sustainable and smart future. Since 1927, we've been known for our commitment to safety and for creating innovative cars that make life less complicated. In 2010, we decided to radically redefine our business, resulting in a totally new generation of cars and technology, as well as steady growth and record sales - and we're only getting started. By 2025, we aim to have yearly sales of 1.2 million cars, 50% of which will be electrical vehicles and sold directly to customers through digital channels. We also intend to strengthen our position as a modern, progressive employer committed to diversity and gender and racial equality. We believe in the power of people. Join us and be part of Volvo Cars' transformation.