Počet záznamů: 1
Adaptive RLS Algorithms Reference Implementations
- 1.0480100 - ÚTIA 2018 RIV CZ eng L4 - Software
Kadlec, Jiří - Likhonina, Raissa
Adaptive RLS Algorithms Reference Implementations.
Interní kód: dsp_1_6 ; 2017
Technické parametry: Algoritmy RLS s exponenciálním nebo směrovým zapomínáním implementované ve čtyřech aritmetikách s různou přesností a spotřebou.
Ekonomické parametry: Software porovnává speciální implementace adaptivních algoritmů identifikace stochastických systémů. Jde o implementace v dvojnásobné přesnosti, jednoduché přesnosti, logaritmické aritmetice 19 a 32bit a celočíselné aritmetice 14bit.
Grant CEP: GA MŠMT(CZ) 8A17006
Klíčová slova: RLS Algorithms * adaptive identification * logarithmic * fixed-point and single-precision floating-point arithmetics
Obor OECD: Computer hardware and architecture
http://sp.utia.cz/index.php?ids=results&id=dsp_1_6
This software presents set of adaptive recursive least squares system identification algorithms based on the Bayesian extensions of real-time adaptive system identification as well as extending the existing recursive least square adaptive algorithms for estimation of time varying parameters in the applications of acoustic signal processing. The included reference adaptive algorithms are implemented in Matlab 2016b. Algorithms serve as „golden“ reference models for the embedded implementations on dedicated processors like Arm Cortex A9 and the FPGA programmable logic accelerators in devices the Xilinx Zynq. Algorithms are numerically robust. Algorithms are implemented in double precision floating point (64bit), single precision floating point (32bit) and in logarithmic arithmetic with precision 32bit and 19bit. This software also presents adaptive recursive least squares system identification algorithms taking advantage of dynamic normalization of the core of the algorithm into the guarantied range 1-, 1 for all variables. These algorithmic cores are suitable for the fixed-point implementation (14bit).
Trvalý link: http://hdl.handle.net/11104/0276767
Počet záznamů: 1