TU Berlin

Fachgebiet Elektronische Mess- und DiagnosetechnikPublikationen

Inhalt des Dokuments

zur Navigation

Publikationen des Fachgebiets MDT

Neural Networks Versus Conventional Filters for Inertial-Sensor-based Attitude Estimation
Zitatschlüssel weber_neural_2020
Autor Weber, Daniel and Gühmann, Clemens and Seel, Thomas
Buchtitel 23rd International Conference on Information Fusion
Jahr 2020
Zusammenfassung Inertial measurement units are commonly used to estimate the attitude of moving objects. Numerous nonlinear filter approaches have been proposed for solving the inherent sensor fusion problem. However, when a large range of different dynamic and static rotational and translational motions is considered, the attainable accuracy is limited by the need for situation-dependent adjustment of accelerometer and gyroscope fusion weights. We investigate to what extent these limitations can be overcome by means of artificial neural networks and how much domain-specific optimization of the neural network model is required to outperform the conventional filter solution. A diverse set of motion recordings with a marker-based optical ground truth is used for performance evaluation and comparison. The proposed neural networks are found to outperform the conventional filter across all motions only if domain-specific optimizations are introduced. We conclude that they are a promising tool for inertial-sensor-based real-time attitude estimation, but both expert knowledge and rich datasets are required to achieve top performance.
Link zur Publikation Download Bibtex Eintrag

Navigation

Direktzugang

Schnellnavigation zur Seite über Nummerneingabe