Options
2003
Conference Paper
Titel
Optimised analysis of low cost inertial measurement-units as a result of Kalman filter
Abstract
In the last few years the development of smaller inertial sensors has been pushed forward. The dimensions of the newest sensors, which are partly still in pilot production measure a few millimetres only. The price of the sensors decreases in line with the size caused in the massive increasing production figures and the application of low cost micro mechanical silicon structures. This makes applications with low cost systems possible in the meantime. The Fraunhofer Technology Development Group (TEG) in Stuttgart has shown the possibility of motion-tracing an object with new low cost miniaturised sensors. The main drawback of low cost IMUs is the sensor drift of the sensors, because with decreasing volume and price the accuracy of the sensors goes down. This disadvantage has obviated the advantage of the low cost sensors up to now. To realise IMUs in spite of these boundary conditions innovative and complex algorithms are needed to compensate for the inaccuracies. On that occasion the calibration of the sensors is an important factor. The research focus includes first of all the handling and analysis of the recorded sensor signals. At this the needed algorithms transcend the algebraical coherence. Complex algorithms from statistical signal processing and the theory of probabilities come into operation. Furthermore the application specific algorithms analyse the custom-designed environment to extract information to improve the accuracy of our orientation tracking. The motivation in scope of developing IMUs for the Fraunhofer TEG is both the need of IMUs in handheld-devices for engineering applications, and for low cost systems for the mass market. Marketable products like computer mice and electronic pens are only two of many applications of IMUs in the low cost segment.
Konferenz
Language
English