Learning control applied to a digital-to-analogue converter
Chapter
Accepted version
Permanent lenke
https://hdl.handle.net/11250/3131241Utgivelsesdato
2023Metadata
Vis full innførselSamlinger
Originalversjon
van Rijt, N., Faza, A., Oomen, T., & Eielsen, A. A. (2023, August). Learning control applied to a digital-to-analogue converter. In 2023 IEEE Conference on Control Technology and Applications (CCTA) (pp. 91-96). IEEE. 10.1109/CCTA54093.2023.10252330Sammendrag
Digital-to-analogue converters (DACs) exhibit several non-ideal effects that deteriorate performance. Methods in feedback control can reduce such effects. Due to implementation limitations, the feedback signal in existing schemes is produced by open-loop observers, known as ΔΣ-modulation, that mitigate the observed adverse effects only partially. Measurement feedback can compensate for non-ideal behaviour and disturbances that are difficult to model. Learning control (LC) is introduced to overcome practical problems of measurement feedback in DACs, therewith omitting the need for accurate open-loop observers. Experimental results demonstrate a 95% improvement in RMS error when using LC with measurement feedback, compared to ΔΣ-modulation using observer feedback.