Master theses / Diploma theses / Compulsory internship

Data acquisition in large industrial vessels such as biogas fermenters or wastewater treatment plants is limited to local measurement points due to limited access to the vessel and the non-transparency of the fluid. To optimize these kinds of plants, the three-dimensional flow field and the spatial distribution of fluid properties such as temperature and electrical conductivity inside the vessel must be known. This can be achieved by the autonomous flow-following sensor particles developed by the HZDR. Equipped with a pressure sensor, an accelerometer, two gyroscopes and a magnetometer, the sensor particle can track the movement inside the vessels and derive the flow field from that. Additionally, the sensor particle gets position information by an ultra-wide-band based localization module (like GPS) as soon as it is on the fluid surface. The motion of the sensor particle is currently tracked with an error-state Kalman filter and yields a reliable tracking of the velocity and position, respectively. However, the tracking time is limited by the propagation of uncertainties of the inertial sensors through the filter. The objective of this master thesis is to extend this tracking time by the use of more advanced tracking algorithms like particle filter or other types of Kalman filters. This includes the following tasks:

• Literature review of advanced filters for motion tracking
• Theoretical comparison and implementing the most promising algorithm in Python
• Verification and performance analysis based on experimental data

• Studies in the area of electrical, mechatronic, mechanical engineering or similar
• Basics of measurement uncertainty, digital signal processing
• Data analysis in Python
• Independent and structured way of working

• Start possible at any time
• Duration according to the respective study regulations