Fuzzy system for assessing bovine fertility according to semen characteristics

• Autor(a) principal: Maziero, Luana Possari [UNESP]
• Data de Publicação: 2022
• Outros Autores: Chacur, Marcelo George Mungai [UNESP], Cremasco, Camila Pires [UNESP], Putti, Fernando Ferrari [UNESP], Gabriel Filho, Luís Roberto Almeida [UNESP]
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://dx.doi.org/10.1016/j.livsci.2022.104821; http://hdl.handle.net/11449/223207
• topic: Animal reproduction; Artificial intelligence; Decision-making tool; Fertility; Mathematical model

In order to maintain the competitiveness of Brazilian livestock, producers are looking for tools to help in the decision-making process in favor of enhancing production management, improving the handling of livestock, and reducing costs. Therefore, the objective of the present work was to elaborate a mathematical model based on fuzzy logic that would be able to provide a bovine fertility score, based on the evaluation of animal semen characteristics. For the development of the variables of the model, the limits established by the Brazilian College of Animal Reproduction were considered for, in conjunction with the Brazilian Ministry of Agriculture, Livestock and Supply. These variables were denominated: Vortex, Motility, Potency, Major Defects, Minor Defects and Total Defects of the semen. Using an ‘If…Then’ rule-base, through the Mamdani inference method, the variables were combined, totaling 735 rules, which provided the Fuzzy Fertility output variable. Thus, there is the classification of unfit oxen, for the ones who presented Fertility less than 1, and the classification of fit oxen, divided into 13 groups according to degree, for the ones that presented results greater than or equal to 1. Thus, the model allows the gradual classification of semen samples classified as suitable, indicating which animals are in better conditions, showing the possibility of migration of the classification of animals, through the degrees of membership. The model proved to be efficient in its classification objective, enabling a new tool to aid the livestock producer and manager.