Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Brésil → 2019 → Modelagem neuro-fuzzy : utilização da bagana de carnaúba para otimização da irrigação na cultura do coentro no semiárido brasileiro

Universidade Estadual Paulista (UNESP) 2019

Modelagem neuro-fuzzy : utilização da bagana de carnaúba para otimização da irrigação na cultura do coentro no semiárido brasileiro

Rodrigueiro, Golbery Rudolf Oliveira

Titre : Modelagem neuro-fuzzy : utilização da bagana de carnaúba para otimização da irrigação na cultura do coentro no semiárido brasileiro

Neuro-fuzzy modeling : use of carnauba bagana to optimize irrigation in coriander culture in the Brazilian semiarid region

Auteur : Rodrigueiro, Golbery Rudolf Oliveira

Université de soutenance : Universidade Estadual Paulista (UNESP)

Grade : Mestre em Agronomia (Irrigação e Drenagem) 2019

Résumé
The use of mulch in the soil to decrease evapotranspiration, temperature, and quantity applied to irrigation depths is the subject of studies and an extensively recommended practice, particularly in semi-arid regions, contributing to the improvement of crop performance and increased retention of soil moisture. In this sense, the reduction of water consumption by crops, without these having reductions in their production, motivated this study, which evaluated the use of carnauba bagana as an option for vegetation cover to reduce the evapotranspiration of coriander (Coriandrum sativum L.). The experiment was carried out in an area belonging to CCEP (Cooperative Cell Education Program) located in the Cipó community, in the municipality of Pentecoste, State of Ceará. , with five primary treatments and five secondary treatments, arranged in subplots submitted to different irrigation depths (50% ; 75% ; 100% ; 125% and 150%) based on crop evapotranspiration (ETc) and different amounts of mulch using the carnauba bagana. The experiment took place in two cycles. To evaluate the productivity of the coriander culture, mathematical models were developed with the adaptive neuro-fuzzy inference system (ANFIS), considering two input variables (irrigation blades and cover levels) and eight output variables (dry mass of the part aerial ; dry root mass ; number of leaves ; plant height) implemented in MatLab software in the Applied and Computational Mathematics Laboratory (LabMAC) belonging to the UNESP Faculty of Science and Engineering (Campus Tupã). In the first cycle, the best results occurred close to 55% of the irrigation blade combined with a range of 40% to 50% of mulch levels, in the second cycle, water consumption shows more satisfactory results ranging from 50% to 80% of the blades used, it is observed that mulch provides productive gain for most in the agronomic characteristics evaluated for the culture of organic coriander, this technique being extremely important.

Présentation

Version intégrale

Page publiée le 18 novembre 2020