Prediction of Fire Spread Based on Geographical Cellular Automata

Abstract

China is a country with frequent forest fires. Forest fires not only damage the forest ecosystem, cause the loss of forest resources, but also threaten the safety of people's lives and property. If the spread trend of forest fire can be predicted quickly and accurately after the fire, it will be of great significance and role to put out the fire and reduce the loss and casualties. In this study, based on the in-depth comparison of domestic and foreign commonly used forest fire spread models and related algorithms, the geographic cellular automata is selected as the diffusion algorithm, combined with the revised Wang Zhengfei fire spread model, and the host development method is adopted to achieve the forest fire spread prediction. Through the comparative analysis of the prediction results in the actual cases, the error is within the acceptable range, achieving the expected effect, meeting the requirements of real-time and efficient simulation of forest fire spread, and providing reliable decision support for the management.