Design of Digital FIR Filters Using Differential Evolution Algorithm

Abstract

The differential evolution (DE) algorithm is a new heuristic approach with three main advantages: it finds the true global minimum of a multimodal search space regardless of the initial parameter values, it has fast convergence, and it uses only a few control parameters. The DE algorithm, which has been proposed particularly for numeric optimization problems, is a population-based algorithm like the genetic algorithms and uses similar operators: crossover, mutation, and selection. In this work, the DE algorithm has been applied to the design of digital finite impulse response filters, and its performance has been compared to that of the genetic algorithm and least squares method.

Robust impaired speech segmentation using neural network mixture model

“…The derivatives obtained directly from the formants frequencies are quite noisy, to filter the noise, the derivatives were obtained via polynomial approximation. A second order polynomial is fit into the trajectory of each of the three formants, the derivative is obtained using (8). To set the decision metric, the absolute sum of the formant derivative is normalized, and the weighted mean and standard deviation were used to set the threshold as depicted by (9) …”

“…One of the first application domains for DE has been signal processing and more specifically the design of a digital filter, see [5]. Many other studies have shown the efficiency of this algorithmic structure in handling these problems, as shown in [6], [7], and [8]. A comparative study focussing on the capability of DE schemes of handling non-standard filter design problems has been presented in [9].…”

Design of Optimized Transition Width Linear Phase FIR Filter using PSO Algorithm with Constriction Factor Approach

“…Evolutionary optimization techniques as Genetic Algorithm [8], Simulated Annealing (SA) [9], Tabu Search [10], Artificial Bee Colony optimization [11], Differential Evolution [12] are implemented for the design of optimal digital filters and proved to be quite efficient by providing better control of performance parameters in addition to high stopband attenuation. Heuristic optimization technique of Genetic Algorithm (GA) gives efficient results for local optimum but is not very successful in determining global optimum and is also considerably slow.…”

Differential evolution schemes for speech segmentation: A comparative study

“…One of the first application domains for DE has been signal processing and more specifically the design of a digital filter, see [5]. Many other studies have shown the efficiency of this algorithmic structure in handling these problems, as shown in [6], [7], and [8]. A comparative study focussing on the capability of DE schemes of handling non-standard filter design problems has been presented in [9].…”