Show all publications

Towards a Multi-Agent Model to Prevent Damage Caused by Cocoa Mirids to Cocoa Pods

Download PDFDownload Bibliography in Open DocumentDownload Bibliography in HTMLDownload BibTeXDownload RISDownload Bibliographical Ontology (RDF)
In Proc. of Les 17e Rencontres des Jeunes Chercheurs en Intelligence Artificielle (RJCIA), Toulouse, France, pp. 10-17, Toulouse, France, 2019.
Agriculture in general and cocoa farm in particular is one of the primary sources of income for several African countries. Although quite useful for the development of Africa, cocoa production is faced with several challenges like the diseases of cocoa trees. Mirid (Sahlbergella singularis) is the worst pest affecting cocoa production across west Africa causing 30% to 40% of damage on cocoa production. It is difficult to control the mirids’ population and its impact on the cacao production. Related works on mirids studied the mirids population as a macroscopic entity without considering local interactions of individual mirids. These models did not consider the behaviours and the environment of mirids which are essential due to their complexity. To overcome these drawbacks, this paper proposes an agent-based model of mirids considering the individual interactions of mirids within the environment. The proposed model consists of agents representing the mirids through their life cycle in cocoa farm environments. It is based on the biological and ecological partial knowledge found in the literature, and is built according to the GAIA methodology. The main assets of this model are the reproduction of the dynamics of the mirids on a cocoa pod and damage caused by them at the micro level for human users (entomologists, agricultural engineers, farmers and decision makers) to thoroughly understand the phenomena and the emergent properties outcomes.
Mirid, Agent-based modeling, Cocoa, GAIA methodology.
Publication Category:
National conference with proceedings
Copyright 2010-2019 © Laboratoire Connaissance et Intelligence Artificielle Distribuées - Université Bourgogne Franche-Comté - Privacy policy