Numerical response of the magnetic permeability as a funcion of the frecuency of NiZn ferrites using Genetic Algorithm

TL;DR

This paper models the frequency-dependent magnetic permeability of NiZn ferrites doped with Yttrium using a genetic algorithm to optimize the model parameters, providing insights for designing magnetic devices.

Contribution

It introduces a novel application of genetic algorithms to fit the magnetic permeability model of Yttrium-doped NiZn ferrites as a function of frequency.

Findings

Genetic Algorithm effectively fits the permeability model.

Model accurately predicts permeability across frequency range.

Provides a basis for designing ferrite-based devices.

Abstract

The magnetic permeability of a ferrite is an important factor in designing devices such as inductors, transformers, and microwave absorbing materials among others. Due to this, it is advisable to study the magnetic permeability of a ferrite as a function of frequency. When an excitation that corresponds to a harmonic magnetic field \textbf{H} is applied to the system, this system responds with a magnetic flux density \textbf{B}; the relation between these two vectors can be expressed as \textbf{B}=$\mu(\omega)$ \textbf{H} . Where $\mu$ is the magnetic permeability. In this paper, ferrites were considered linear, homogeneous, and isotropic materials. A magnetic permeability model was applied to NiZn ferrites doped with Yttrium. The parameters of the model were adjusted using the Genetic Algorithm. In the computer science field of artificial intelligence, Genetic Algorithms and Machine Learning does rely upon nature's bounty for both inspiration nature's and mechanisms. Genetic Algorithms are probabilistic search procedures which generate solutions to optimization problems using techniques inspired by natural evolution, such as inheritance, mutation, selection, and crossover. For the numerical fitting usually is used a nonlinear least square method, this algorithm is based on calculus by starting from an initial set of variable values. This approach is mathematically elegant compared to the exhaustive or random searches but tends easily to get stuck in local minima. On the other hand, random methods use some probabilistic calculations to find variable sets. They tend to be slower but have greater success at finding the global minimum regardless of the initial values of the variables