Evaluation and comparison the results of optimization of forest fire spreading model based on cellular automata using PSO and ABC algorithms

Forest fire is one of the most common ecological disaster which it''s accurate spread prediction is a very essential issue in minimizing it''s destructive effects. This phenomenon depends on factors such as topography، vegetation and climate. Among the available models، experimental deterministic models that are presented in the form of raster such as cellular automata are more popular due to their simplicity and ability in modeling complex systems. Many simulator systems have been developed in order to simulate and predict fire spread by cellular automata. The quality of their results depends on accuracy and reliability of input parameters as well as model complexity which most of these parameters have a degree of uncertainty. A constructive suggestion to overcome the problem of uncertainty is using a two phase simulation approach. In this approach، at first، model parameters will be optimized by comparing simulation results with reality، then simulation model will be used for simulating next step of fire spread considering optimal values which already has been obtained in the first step. One of the important points in designing this system is using optimal optimization method. In this study، two optimization methods named Particle Swarm Optimization (PSO) and Artificial Bee Colony (ABC) have been used to overcome the problem of uncertainty and enhance the precision of forest fire spread modeling and performe two phase simulation approach for part forests of Gilan provinces. Results show that Artificial Bee Colony has higher ability than Particle Swarm Optimization in producing optimal parameters of the model.