| Fire burning, characteristics of smoke movement, the size and development tendency of smoke signals all can be influenced by low pressure and low oxygen concentration environmen directly such as Tibet. The fire detection is based on the concentration, intensity and development tendency of fire products. Therefore, this can have an effect on fire detections. The types of fire detectors and installation site continued to use the standards in normal pressure environment. This situation is lack of scientific proof and data support, so there are high probabilities of false and mistake fire alarms.The design of fire smoke detectors is based on quantitative smoke characteristics express. It is necessary to know the smoke signals quantitative laws in low pressure and low oxygen concentration environment in a confined ceiling jet zone to guide the fire detectors design, selection and installation to satisfy the fire prevention demands.In order to study the characteristics of smoke temperature in the confined ceiling jet zone, the theoretical analysis and experimental validation have been done. Firstly , the steady heat release rate fire smoke temperatures in a ceiling jet zone are analyzed in different atmospheric pressure conditions. The theoretical analysis is based on pressure similar model and ceiling jet theory. The experiments were done in Hefei (about 1 atmospheric pressure) and Lhasa (about 0.67 atmospheric pressure) in order to validate the theoretical analysis. The ethanol fire (43.5cm×43.5cm) is chosen as the test fire in the ISO fire detection room. The results show that the curves can fit well between theoretical prediction and experimental test.Aiming at the deviation between the experiment results and the engineering experiment formulation, the smoke plumes and ceiling jet theory are used to deduced the smoke density in ceiling jet zone. The results show that smoke density rises up with the increase of combustion heat release rate. If the heat release rate is the same, the smoke density also can be the same. The convection equations are used to realize the function simulation in the process of discussing the sensitivities of photoelectric and ionization smoke detectors in two kinds of different environment. The theoretical analysis shows the heat convection in low pressure normal environment and in normal pressure low gravity can get similar results. So the researched in microgravity are used to analyze the smoke detector sensitivity in low pressure ambient. The results show the semi-quantitative theoretical analysis can explain the experiment results well.Smoke detectors can not get the satisfying alarm to flammable fire in Tibet. Firstly,This paper presents a synthesis method of GM (1,1) grey prediction model and artificial neural network in advance to detect fire and to make it work in the environment. The fire detection algorithm can give the accurate fire probabilities. However, some of them are so fuzzy that it is can not give us the fire or non-fire decision. Then this paper presents a synthesis method of GM (1,1) grey prediction model and adaptive neuro-fuzzy inference system (ANFIS) in advance to detect fire and to make it work in the environment. In this process, the collected data are filtered by adaptive threshold and tendency, then the data are input the grey GM(1,1) predict model to anticipate the development of fire signals without any assumption, meanwhile, check out the accuracy of prediction. Then the data are input the ANFIS whose initial membership functions were trained by experiment data in Tibet. This work will supply useful suggestions with the fire detection design in low ambient pressure and low oxygen concentration such as Tibet etc.
|
PDF Full Text Request