Cadeia de Markov aplicada ao manejo de pragas em lavoura cafeeira

Abstract

Regarding the nature of systemsr behavior, it can be classified into two types: deterministic and probabilistic ones. Deterministic systems have a known behavior or determinable over time, which can not be said of probabilistics. These are based on event probabilities and require an auxiliary tool to be analysed. A tool that can assist in its analysis is the Markov chain, the object study of this work. Through a management problem coffee crop pests a Markov chain model was proposed, according to a methodology developed. We addressed two dimensions, namely, impact and spraying. The results of three consecutive years were raised by addressing percentiles of recurrent transition to the state without incidence and without spraying, as well as the percentage of states without incidence and without spraying. Also the sectors with maximum and minimum impact and high and low spraying for the time periods were raised. The Markov chain proved an effective tool in pests monitoring and the system global analysis, which is essential for the development of a precision agriculture