Bachelor or Master Student Electrical, Mechanical Engineering, Basic Sciences or similar (f/m/x) - Development and refining of a neural network

Location

Jülich | Germany

Job description

Bookmark

Course paper / final thesis, Student placement, Internship

Development and refining of a neural network for the bias correction between simulation and measurement technologies using image processing

Solar power plant Jülich

Starting date

1 April 2024

Duration of contract

up to 6 months

Remuneration

up to the German TVöD 5

Type of employment

Full-time

It is necessary to develop a Sun flux density measurement (FDM) method for every single type of receiver in concentrated solar power (CSP) plants.

The proposed procedure here is an AI-assisted digital twin using image processing tools as groundtruth. This model presents a hybrid scope between a data-driven model for prediction and a conventional method for indirect measurement of the receiver.

This Bachelor’s End Thesis or Master’s End Thesis offer is focused on the Work Package 2 (WP 2) of the project (see below) and consists of the creation of a neural network (NN) for plane image mapping in the plane image space for prediction of flux density in a white target given a plant and atmospheric conditions. The neural network will be constituted from pre-existing technologies for image generation (UNet, GANs…) and the training of this network will be divided in different stages: pre-training for different simulation conditions and training for real conditions against real data.

WP1: development of a digital twin of the power plant based on its data platform for real-time measurements in the aperture plane of the cavity or in open plane receivers

WP2: development and refining of a neural network for the bias correction between simulation and measurement technologies using image processing

WP3: integration of the cavity in the simulation for FDM inside and bias control using the radiation spillage in the shield of the cavity and the NN developed also for bias correction in WP2

• Study of Engineering, Electrical Engineering, Mechanical Engineering or Basic Sciences
• Python programming, fluent English, Heat Transfer basics, Neural Networks basics, CSP technologies knowledge (not mandatory, but valuable), digital image processing, teamwork and motivation

Look forward to a fulfilling job with an employer who appreciates your commitment and supports your personal and professional development. Our unique infrastructure offers you a working environment in which you have unparalleled scope to develop your creative ideas and accomplish your professional objectives. Our human resources policy places great value on a healthy family and work-life-balance as well as equal opportunities for persons of all genders (f/m/x). Individuals with disabilities will be given preferential consideration in the event their qualifications are equivalent to those of other candidates.

Job tags

Salary