With the development of coal industry, methane explosion accidents happen frequently, which has been an important factor restricting the safe production of coal mine. The existence of methane, oxygen and fire source is the reason causing the methane explosion. Therefore, the application of reliable methane concentration monitoring technology in real-time has important meanings for society and economy in order to ensure the safety of state property and people life.Firstly, the conventional detection methods of methane and the researching status quo of the optical fiber methane sensors at home and abroad have been summarized in this dissertation. Then we chose the optical fiber methane detecting network as the researched object. Based on infrared absorption spectrum theory, a quasi-distributed optical fiber methane sensor network has been designed, which consists of distributed feedback laser diode (DFB LD), optical switch and space division multiple technology, both differential absorption and harmonic detection methods have been used respectively.Secondly, the light source, the sensing, the circuit of photo-electric conversion, the signal collecting and processing modules of the network have been presented separately. The key technologies of methane absorption in the R(4) line of 2V3 band lines with the wavelength of 1650.96nm, mode hopping characteristics of the DFB LD and the design of gas cell with anti-pollution device have been researched specially. In addition, the preamplifier circuit and band-pass filter have been designed. The special virtual monitoring system of methane concentration with data acquisition card and LabVIEW programming software has been exploited to implement the waveform visualization and numerical display of methane concentration in real-time, sound and light alarming in time.Finally, the methane measurement and optical fiber loss experiments have been implemented by employing differential absorption and harmonic detection schemes respectively. Repeatability and stability of the network have been verified. The experimental results are consistent with the theoretical analysis. Through analysis, the sensitivity of network is 0.0018uw/ppm, and the resolution can reach 96ppm. |