Evolutionary-Statistical System: a Parallel Method for Improving Forest Fire Spread Prediction

BIANCHINI, GERMÁN; CAYMES SCUTARI, PAOLA; MENDEZ GARABETTI, MIGUEL

Journal of Computational Science

Elsevier

Lugar: Amsterdam; Año: 2015 vol. 6 p. 58 - 66

1877-7503

Fighting fires is a very risky job, where loss of life is a realpossibility. Proper training is essential. Several firemen academies offercourses and programs whose goal is to enhance the ability of fire andemergency services to deal more effectively with fire. Among the toolsthat can be found in the training process are fire simulators, which areused both for training and for the prediction of forest fires. In manycases, the used simulators are based on models that present a series oflimitations related to the need for a large number of input parameters.Moreover, such parameters often have some degree of uncertainty due to theimpossibility of measuring all of them in real time. Therefore, they haveto be estimated from indirect measurements, which negatively impacts onthe output of the model. In this paper we present a method which combinesStatistical Analysis with Parallel Evolutionary Algorithms to improve thequality of the model output.