Designing an Agent Based Model for the Efficient Removal of Red Imported Fire Ant Colonies

摘要

Red imported fire ants have proven to be a bane of the existence of many individuals in the southern United States. Defining methodologies for prevention, removal, and destruction of these invasive insects continues to be difficult due to their highly developed survival instincts. Multiple methods and processes exist to force the destruction of red imported fire ants yet their sheer numbers and ability to return to previously cleared spaces defines a need for greater understanding of the ant-to-ant interaction within these beds. The primary purpose of this research is to design a working model of a red imported fire ant bed that can be improved upon by interactions with the University of Texas at Tyler biology department. To meet this goal an understanding of agent-based models is first needed. Agent based models (ABM) have the ability to generate complex behavior using easily understandable rules that can be altered by an individual with less knowledge of the computing process and more involvement in the system process. Another benefit of these models is their highly parallelized nature leading to other technological research routes to speed up the processes and create more cross discipline projects. Based on the ABM's highly parallelized nature, implementing this model on a system that can harness the low cost parallel processing environments of general purpose graphics processors and multicore CPUs is our ultimate goal. This paper outlines our initial steps in defining the model and designing a system that works effectively at a small scale on a sequential environment. Our future plans to port this model to a parallel processing environment are also detailed.