aktuelle Themen
Auf dieser Seite finden Sie die aktuell formulierten Themen für studentische Arbeiten. Darüber hinaus besteht stets die Möglichkeit, entsprechend Ihren Interessen individuelle Themen für Studien- und Diplomarbeiten zu vereinbaren. Bitte informieren Sie sich dazu in unserer Liste der Themengebiete und Ansprechpartner für studentische Arbeiten.
Accurate Position Estimation by Exploiting an Inertial Measurement Unit
- Betreuer: Dr.-Ing. Christian Scheunert / Dr.-Ing. Merve Sefunç
- Art der Arbeit: Diplomarbeit
Position estimation of a device or an object such as aircraft, humans and robots, relative to a reference frame is challenging yet necessary task for many applications. Precision gyroscopes, ring laser for example, are too expensive and bulky for most applications and so less accurate MEMS (Micro Electrical Mechanical System) devices are used in a majority of applications. The use of relatively cheap sensors is important from a practical point of view; however, low-cost sensors seldom provide good performance due to measurement inaccuracies in various environments. Inertial based sensor methods, also known as inertial measurement units (IMU), are comprised of sensors such as accelerometers, gyroscopes and magnetometers. IMUs offer good signals with high rate during fast motions but are sensitive to accumulated drift due to double integration during estimation of position. The importance of using these sensors is primarily to determine the position and orientation of a particular device and/or object. An accelerometer as a sensor measures the linear acceleration, of which velocity is determined from it if integrated once; for position, integration is done twice. Results produced only by an accelerometer have been unsuitable and of poor accuracy due to the fact that they suffer from extensive noise and accumulated drift. This can be compensated for by the use of a gyroscope and magnetometer. Various data fusion techniques have been developed to overcome this serious drift problems caused by the accumulation of measurement errors over long periods. Therefore, the fusion of an accelerometer, gyroscope and magnetometer sensor is suitable to determine the pose of an object and to make up for the weakness of one over the other. In order to ensure an accurate position estimation by exploiting the data collected by IMU, a low complexity sensor fusion algorithm is to be implemented. The following tasks are to be processed within the scope of the work: - Literature review on the position estimation with special focus on the low complexity and computationally efficient sensor fusion algorithms - Implementing an algorithm on Python or C++ to estimate the position of the object that IMU is attached - Combining the sensor fusion algorithm with the calibration algorithm (which will be provided) to get accurate position information of the sensor unit Requirements: - Good knowledge in digital signal processing and system theory - Programming skills in Python or C++
Implementation of an Online Sensor Calibration Algorithm by Exploiting Machine Learning for an Inertial Measurement Unit
- Betreuer: Dr.-Ing. Christian Scheunert / Dr.-Ing. Merve Sefunç
- Art der Arbeit: Diplomarbeit
The accurate measurement of orientation plays a critical role in a range of fields including: aerospace, robotics, navigation and human motion analysis and machine interaction. Whilst a variety of technologies enable the measurement of orientation, inertial based sensory systems have the advantage of being completely self contained such that the measurement entity is constrained neither in motion nor to any specific environment or location. An IMU (Inertial Measurement Unit) consists of gyroscopes, accelerometers and sometimes magnetometers enabling the tracking of rotational and translational movements. In order to measure in three dimensions, tri-axis sensors consisting of 3 mutually orthogonal sensitive axes are required.
Micro Electrical Mechanical System (MEMS) devices are used in a majority of applications due to their simplicity and low cost. However, they tend to have lower accuracy and require calibration to improve the performance by removing structural errors in the sensor outputs. Even two sensors from the same manufacturer production run may yield slightly different readings due to the sample to sample manufacturing variations. Two different sensors may respond differently in similar conditions because of the differences in sensor design. Sensors subject to heat, cold, shock, humidity etc. during storage, shipment and/or assembly may show a change in response. Some sensor technologies age and their response will naturally change over time - requiring periodic re-calibration. In order to achieve the best possible accuracy, a sensor should be calibrated in the system where it will be used.
In order to ensure the accurate readings from an IMU, an online machine learning based calibration algorithm is to be implemented. The following tasks are to be processed within the scope of the work:
- Literature review on the sensor calibration with special focus on the machine learning approaches
- Implementing an algorithm on Python to calibrate an IMU that operates on a Raspberry Pi computer in order to provide accurate sensor readings
- Combine the calibration algorithm with a sensor fusion algorithm (which will be provided) to get accurate orientation of the sensor unit
- Compare the performance of the implemented algorithm with the conventional sensor calibration
- Good knowledge in digital signal processing and system theory
- Programming skills in Python or C++