Development of a Magnetic Absolute Encoder Using Eccentric Structure and Long Short-Term Memory
Akishi Takeyama, Shota Komatsuzaki, Keita Sado, Yuki Nagastu, Hideki Hashimoto
- Year
- 2021
- Citations
- 2
Abstract
Motor is important device that support the foundation of our lives. For high-precision control of robots, it is necessary to improve the accuracy of the encoder, which is a position sensor. High resolution can be achieved by using multi-pole magnets. However, the absolute angle is lost. In the previous research, eccentric rotation was used to achieve high resolution without loss of absolute angle. But the number of sensors to be used increases for higher resolution by further increasing the number of poles. When the number of sensors is reduced in relation to the number of magnetic poles of the magnet, a large angle error will occur. In this paper, we proposed a method to calculate the angle when the number of sensors is reduced by using Long Short-Term Memory. The proposed method reduces the large angle error when the number of sensors is reduced.
Keywords
Related papers
Statistical Learning Theory
Yuhai Wu, Vladimir Vapnik
1999
Artificial intelligence: a modern approach
1995
Fractional Differential Equations
Igor Podlubný
2025
Applied Nonlinear Control
Jean-Jacques Slotine, Weiping Li
1991