Situation Prediction of Fire Management System Based on BP Neural Network

Abstract

This paper introduces the principle of BP Neural Network to achieve Training algorithm, which work with Hopfield neural network associative memory to achieve prediction fire in the semi-closed space of the community. The BP Neural Network is regarded as a nonlinear mapping from input to output. Based on the BP neural network algorithm by the software monitoring technology obtain the prediction model which predict the output value is closed to the real value The high effectiveness of Artificial Neural Network is verified by the comparison of specific field simulation. Once the fire happened, People can get fire of extinguishing materials in time by the color of the smoke and fire emitted to prevent the expansion of the fire. Consequently, the fire disasters can be predicted and prevented through the pattern of the fire model.