D3-2.2 - Towards Hardware Implementation of Homogeneous Multi-Sensor Self-X System through Simulation-Based Virtual Sensor Modeling

Event

23. ITG/GMA-Fachtagung Sensoren und Messsysteme 2026
2026-06-09 - 2026-06-10
Nürnberg

Band

Vorträge

Chapter

Messunsicherheit und Selbstvalidierung

Author(s)

E. Gerken, A. König - University of Kaiserslautern-Landau, Kaiserslautern

Pages

264 - 269

DOI

10.5162/sensoren2026/D3-2.2

ISBN

978-3-910600-11-9

Price

free

Abstract

Modern sensor systems are expected to provide reliable and consistent measurement quality even under non-ideal mechanical, environmental, or operational conditions. In many applications, achieving sustained high measurement quality with a single sensor depends on a high-quality sensing element, precise mechanical integration, and application-specific calibration or compensation, which increases system cost and design complexity. This work investigates a Self-X sensor system in which several low-cost and imperfect homogeneous sensors are used instead of relying on a single high-quality and precisely mounted sensor. The main objective is to improve and stabilize the estimation of the target measurand through algorithmic fusion of redundant measurement channels. While previous work mainly focused on testing this hypothesis using simulation-based models, the present study represents a first step toward hardware realization. For this purpose, measurements from one real tunnel magnetoresistance (TMR) sensor were acquired at three different positions around a rotating magnet and used to emulate a homogeneous multi-sensor configuration. After alignment and resampling, the resulting channels were combined using dimensionality reduction to generate a virtual sensor representation. Compared with the best individual sensor, the virtual sensor reduced mean absolute error (MAE) from 2.214° to 0.953° and root mean square error (RMSE) from 2.854° to 1.288°. This corresponds to improvements of approximately 57% and 55%, respectively. The results indicate that multi-sensor fusion can compensate for position-dependent distortions and provide a useful basis for further hardware development of the Self-X sensor concept. Future work will focus on moving towards a simultaneous multi-channel hardware configuration, including configurations with more than three sensors and extended measurement campaigns under various operating conditions.