Genetic Algorithm: Application in the Decision Support System for Selecting a Machine and Tractor Fleet

Abstract

The considered problem is to select a machine and tractor fleet to perform an annual cycle of work in agribusiness. There are many criteria for selecting a machine-tractor fleet and mathematical models based on them to solve this task. The complexity of the problem lies in many numerous solutions. A specialist carries out the assessment and final choice of machine and tractor fleet, so the number of variants offered for consideration should not exceed a reasonable number. In this work, the task is to reduce the number of solutions for selecting a machine and tractor fleet. To solve this problem a combined application of two approaches is proposed. The first one is based on choosing the length of periods that make up the annual work cycle. The second is based on the application of a genetic algorithm based on John Holland's evolutionary algorithm and the selection hypothesis. We held the adaptation of the genetic algorithm and presented the result of its testing. It demonstrates the effectiveness of this approach. It is concluded that the application of the genetic algorithm to the task of reducing the number of solutions allows to do it as well as to cut the time spent on performing calculations for selecting a machine and tractor fleet for cultivation of grain crops.