|
Accession number;99A0938093
|
| Title;A Study on Adaptive Digital Filters Based on Evolutionary Computation. |
| Author;
ABE MASAHIDE
(Tohoku Univ.)
|
Journal Title;Record of Electrical and Communication Engineering Conversazione, Tohoku University
|
Journal Code:F0511A
|
ISSN:0385-7719
|
|
VOL.68;NO.1;PAGE.108-110(1999)
|
| Figure&Table&Reference;FIG.2, REF.7 |
| Pub. Country;Japan |
| Language;Japanese |
| Abstract;This paper proposes evolutionary digital filters(EDFs) as new adaptive digital filters based on evolutionary computation. The adaptive algorithm of EDFs controls and changes the coefficients of inner digital filters using the cloning method or the mating method. This adaptive algorithm of EDFs is of a non-gradient and multi-point search type. Thus, the EDF has following advantages: (1) the adaptive algorithm of the EDF is not susceptible to local minimum problems that arise from a multiple-peak surface, (2) the EDF can adopt the various error functions as the fitness function according to application, (3) the adaptive algorithm of the EDF has a self-stabilizing feature. The EDF has been implemented on parallel processors in order to reduce the total execution time. The partitioned EDF on four DSPs takes 7.15msec per output sample to execute the adaptive filtering, and requires 1/3.97 times as long execution time as the EDF on the single DSP does. (author abst.) |
|
|
|
Related Articles;
|
