Doctorado en Ingeniería (tesis)
URI permanente para esta colección
Examinar
Examinando Doctorado en Ingeniería (tesis) por Materia "Agricultura de precisión"
Mostrando 1 - 1 de 1
Resultados por página
Opciones de ordenación
Ítem Smart insect-pest management for cotton crops(Universidad EAFIT, 2024) Toscano Miranda, Raúl Emiro; Aguilar Castro, José; Toro Bermúdez, Muricio; Caro Piñerez, ManuelIn this research, we address the problem of smart insect-pest management for cotton crops. For the study of this problem, we have positioned it in the framework of the paradigm of Smart agriculture. In this context, Smart agriculture, also known as precision agriculture or digital agriculture, involves the use of advanced technologies to improve agricultural productivity, efficiency, and sustainability. Its focus is to use data-driven and innovative approaches to optimize farming practices and reduce resource waste while ensuring food security. The development of approaches to aid in decision-making for smart insect-pest management for agriculture is necessary to avoid the massive spread of insect pests and the increase in environmental impact. Despite the existence of advances in smart agriculture, integrated management of insect pests remains a challenge. To address this problem, our objective was to develop methodologies, models, and approaches to support decision-making related to smart insect-pest management for cotton crops. To achieve this objective, several sub-objectives were raised, the first one was to design a metacognitive architecture for the smart management of cotton pests, the second was to implement knowledge models for the smart management of cotton pests, and the third was to implement novel AI concepts for the development of knowledge models. Particularly, several research articles were developed to meet the objectives proposed in this thesis. Initially, a review article on the latest trends in Smart agriculture using artificial intelligence and sensing techniques for the management of insect pests and diseases in cotton was carried out. On the other hand, for the first sub-objective, an article was conducted where a metacognitive architecture with metacognitive tasks (meta-memory, meta-learning, meta-reasoning, meta-comprehension, and meta-knowledge) was proposed for smart-pest management of cotton. To meet the second sub-objective, two articles were proposed. The first article is a classification model of the cotton boll-weevil population and the second article presented a fuzzy classification system to analyze the yield of cotton production. Regarding the third sub-objective, two articles were proposed. The first article is about a system with autonomous cycles of data analysis tasks for the integrated management of cotton. And the second article shows how to enhance the insect pest classification in cotton using Transfer Learning techniques. In each article, the strategies/models were evaluated using various datasets. The results showed the capacity of the developed methodologies and models for decision-making in smart insect-pest management for cotton crops. Specifically, our proposals allow the prediction of the boll-weevil behaviors, the diagnosis/prediction of cotton yield, and the prescription of strategies for cotton management into a framework of a meta-cognitive architecture, with good results in performance metrics.