Research On Fire Information Detection Algorithm Based On Fuzzy Neural Network In Intelligent Home Information Fusion

With the progress of science and technology, continuous improvement of people’s living standards, intelligent home life has become a trend. Comfort is no longer the only standard home life, security, intelligence more and more people’s attention. The frequency and scale of modern home fires is growing, causing great loss of life and property, the study of intelligent home fire protection system has very important and practical significance. The primary work of the intelligent home fire detection system that can identify the fire as soon as possible, through the control system or other linkage system, to minimize the damage.This paper studies the complexity of the nonlinear structure fire signals, etc., and then study the theory of fuzzy theory and RBF neural networks were designed fire detection system model and RBF fuzzy neural network, and use Matlab simulation experiments, analysis results, simply relying on a particular algorithm to process and can not get the desired effect.Further proposed fuzzy systems and neural networks, complementary advantages, have devised a six-layer fuzzy neural network structure. Among the fuzzy neural system, the entire fuzzy inference section fuzzy neural network system to complete the original neural network; train the neural network is performed by an error backpropagation method to obtain weights for the neural network fire detection the environment; training be modified after the membership functions and fuzzy rules, as a further reason to use. When an incoming test sample, according to the trained fuzzy neural detection analysis system, able to draw the probability of fires. Further introduction of smoke duration during smoldering fires and open fires can be difficult to judge whether further judge, and can effectively improve the anti-jamming, when fire detection to make more accurate judgments.Simulation results show that the fuzzy neural network can get the desired results prove fuzzy neural network fire detection is reasonable, it is possible to accurately predict, reduce false positives, effectively improve the stability and reliability of fire detection, and has a certain anti-jamming capability...