Research On Smart Home Fire Prevention System Based On Extension Neural Network

Fire prevention system is an important part of the smart home system, and provides an important guarantee for the safety of personal property. Fire prevention system based on smart home platform need to be able to identify the different home fire situation, and to make accurate prediction in the early stages of fire. The traditional fire detectors are based on single parameter for fire detection, and the detection algorithm is simple threshold judgment, trend analysis, etc. There is high probability of false positives and false negatives, and the scope of application has great limitations. In this paper, based on the actual characteristics of the home environment, a fire prevention system based on multi sensor information fusion technology is studied, and the extension neural network is applied to the fire information fusion.Firstly, the research status of fire prevention system, the generation mechanism of the fire and the related technology of fire protection is studied. The technology of multi sensor information fusion is used to design the fire prevention system through the analysis of the characteristics of fire and the characteristics of different fire detectors and there application scope. Secondly, the overall framework of the smart home fire prevention system is designed according to the actual characteristics of the home environment. And temperature, smoke concentration, CO gas concentration are chosen as the fire information fusion parameters, and the realization method of fire information fusion algorithm is also designed. Then, the fire information fusion algorithm is designed, and the extension neural network is introduced into the fire information fusion. The national standard experimental fire data is selected to carry on the network training. In the decision-making level, the smoke duration is introduced as a judgment basis for the fire, and improves the reliability of the system. Finally, the smart home fire system platform is designed, including the central controller, information collection module(temperature acquisition module, smoke concentration acquisition module and CO concentration acquisition module), wireless transmission module, alarm module and the PC monitoring program, and the system is tested to verify its validity and reliability.