Research On The Technology Of Flammable Gas Detection And Alarm

Flammable gas which is a new type of energy has got extensive application in industry and daily life. But because it is flammable and explosive, the flammable gas will also seriously threaten the public's property and life safety when the consistency over the explosion limiter in the air. However, our country have not yet formed a sophisticated system of flammable gases detection and alarming for the historical and recognition reasons. Especially in petrochemical industry, there have been many safety problems existed in the area, such as false alarming and detection failures. Therefore, this paper conducts the research on the technology of flammable gas detection and alarming to meet the petrochemical industry's demand. And the main contributions of this paper are as follows:1. The paper introduced the basic theory and methods of the flammable gases detection. And on this basis, for overcoming the shortcoming of the method that used the single gas sensor to detect single or multi-gas, a new thinking of using the sensor array to get the information of flammable gases and then to distinguish and analysis it using the multi-sensor information fusion technique is proposed in this paper.2. Designed a system to detect the flammable gases of CH4, CO, H2, which widely exists among petrochemical industry. In the detection system, the sensor array was composed by 6 SnO2 series sensors (TGS813, TGS822, TGS203, MQ-8, MQ-7, MQ-4). And this paper has also designed the hardware for the detection system which mainly included power module, gas detector module, delay circuit, LCD display and alarm module.3. And the paper analyzed the advantage of the neural networks that applying in the flammable gas detection, then proposed the method of making the information fusion of the data which got by the gas sensor array using the RBF (Radical Basis Function) neural networks. And it is proved by simulation experiments that the method of the RBF neural networks combine with gas sensor array can be used to qualitative identify and quantitative analysis of single flammable gas and quantitative analysis of mixed flammable gas. And the simulation result verified that this method has a very good effect in the flammable gas detection and also has the desirable application value in practice.4. The paper proposed to construct the detection and alarming system by wireless sensor networks to overcome the shortcomings of the conventional wired alert networks. And considering of the energy saving, this paper proposed a hierarchical routing algorithm: TEEN-PE which is based on TEEN. And it is proved by simulation experiments in NS-2 that TEEN-PE has an advantage over TEEN protocol on energy- saving and has the application value in practice.