The paper presents the results of using a hybrid algorithm in the problem of digital filter design, using its analog prototype as an example. In the parameter determination process, two main stages can be distinguished: the first stage – the application of a steady-state genetic algorithm, and the second – a selected static optimization method, assuming that the results obtained in the first stage will serve as starting conditions for the second stage. This combination ensures the utilization of the advantages of both methods, namely the effectiveness of the genetic algorithm in searching large spaces as well as the good convergence of the classical method in the vicinity of the global minimum point.Digital filters are one of the fundamental elements of digital signal processing. Taking into account the nature of the passband, they can be divided into: high-pass, low-pass, band-stop, and band-pass.
Considering their structure, filters are distinguished into: finite impulse response (FIR), infinite impulse response (IIR), adaptive, nonlinear, etc. [6].
FIR filters (eng. (…)
The article contains 17501 characters.
Source: Logistyka Magazine
Hybrid algorithm in digital filter design – PDF file for download