Optimal Crops Selection using Multiobjective Evolutionary Algorithms

Ricardo Brunelli, Christian von Lücken


Farm managers have to deal with many conflicting objectives when planning which crop to cultivate. Soil characteristics are extremely important when determining yield potential. Fertilization and liming are commonly used to adapt soils to the nutritional requirements of the crops to be cultivated. Planting the crop that will best fit the soil characteristics is an interesting alternative to minimize the need for soil treatment, reducing costs and potential environmental damages. In addition, farmers usually look for investments that offer the greatest potential earnings with the least possible risks. According to the objectives to be considered the crop selection problem may be difficult to solve using traditional tools. Therefore, this work proposes an approach based on Multiobjective Evolutionary Algorithms to help in the selection of an appropriate cultivation plan considering five crop alternatives and five objectives simultaneously.


multiobjective optimization, crops selection, evolutionary algorithms

Full Text:


DOI: https://doi.org/10.1609/aimag.v30i2.2212

Copyright © 2017, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.