Implementation of Genetic Algorithms Into Development of Mechatronic Multi-Edge’s Grinder Design

Abstract

Authors of the study propose the method aiming at determination of design properties of a multi-edge grinder. This method is based on a genetic algorithm and its purpose is to optimise the geometric shape of the cutting edges. The input data include population of individuals. Each individual is represented by a set of cutting disks. Whereas the fitness function was assumed as a combination of several postulates of the mechanical design foundations. Those postulates include mechanical, design and energy metrics. Each individual constitutes a complete solution of the disk set whereas the population represents the entire class of solutions. The fitness function of an individual is calculated as the fitness average of each disk supplemented by information describing the relationship between both disks and discs. The method for calculating function values was selected to ensure its maximisation in the process of evolution. Despite promising results of the genetic algorithms operation, one can consider improvement of the method efficiency. The authors used morphological operations in order to better adjust the method to the task. Results of the simulations were verified in laboratory conditions with positive effects.