Error Analysis And Research On Fire Alarm System For Urban Tunnel

Once the fire is in the urban tunnel, due to the closed, narrow and long special circumstances, it would result in incalculable disaster. Therefore, the fire alarm detection of the urban tunnel is badly important to the fire prevention.Fire alarm system as the indispensable urban tunnel monitoring system for the city has played a major role in transportation safety. Correspondingly, there are various forms of fire alarm system which are applied to many urban tunnels.The fire must meet two basic conditions, combustible materials and oxygen, to raise the alarm of fire alarm system accuracy and reduce error of the alarm. We must understand the ways of fire. Understanding of the causes of the fire is to avoid and reduce the incidence of fire for us in the design of fire alarm system, leading to a minimum rate of fire and the lowest disaster.The key components of fire alarm system are fire detectors, based on the characteristics of fire. There are a lot of fire detectors, according to the principle, which can be divided into distributed temperature cable detector, fiber optic detector, dual-wavelength flame detector separately. Fiber Bragg Grating fire detector is the relatively high accuracy detector. Its principle is that the silica molecular bond will produce lattice vibrations because of the temperature elevation, and result in optical fiber transmission to produce scattering (Raman scattering).These weak changes in optical information are the condition for controller to determine fire. Wuhan Zhongshan Road Tunnel project involved in the article is based on fiber optic detector. For the information processing, there are two options to trigger the alarm, setting the data parameters of the temperature jumping and the temperature threshold. In the designing of Wuhan Zhongshan Road Tunnel fire alarm system, there are 20 fire zones. Each zone at the top of the tunnel lays out 10 Fiber Bragg Grating sensors. The spacing is 5m, by setting the temperature threshold to show the specific area alarm.This paper, based on the Wuhan Zhongshan Road Tunnel, analyzes and studies performance of the fire alarm detector, analyzes the technical principles of fire alarm detector. To find out the reasons of the alarm error and create a condition model which could reduce alarm errors:integrating the results of three factors, which are the temperature jump, the temperature threshold and the adjacent temperature, as the condition whether to issue a fire alarm. In the existing fire alarm system, create a BP neural network simulation model which is based on Wuhan Zhongshan Road Tunnel project, access to actual operation data as the training sample data,. Input the sample data into the simulation model, set expected output as target output, compare and analyze analog output and actual operate result in project, then observe the effectiveness of this strategy, so as to provide support for designing fire alarm. After validating the model presented in this paper, it is proved that the model is a reliable, high accurate fire alarm system.