Optical Fiber Networks And Implementations For Multiple Points And Multiple Gas Sensing

With the continuous development of the Internet of Things technology,the multi-gas multi-point sensing network has become significant in many fields,such as the applications inenergyindustry,environmentalmonitoringand ecology.The electricgas sensingtechnology has advantages of maturity and deversity;there are still some improvementto work with,such as shortlifetime,cross interference,high cost and limiteddetection accuracy.The optical fiber gas sensingtechonology has becomeanemerging gas detection technology.It has severalsuperioritiessuch ashigh accuracy,low cross interference and simplifiednetwork structures.Due to the available and favorableoptical detection techniques,the integrationwithmulti-gas sensing is still indevelopment.To establish an effective multi-gas multi-point optical sensing network has becomeacttractive and necessary.In this work,amulti-gas multi-pointoptical fiber sensingnetwork based on spectral absorption principle isproposed and experiementlly demonstrated.The theoretical analysis and experimental study were carried out in detail.Based on thegas spectral absorption theory,the relationship between gas concentration and spectral absorption was deduced and simulated.Consideringthe obvious absorption spectrum in the optical fiber communication window,methaneand carbon dioxideare characterisated in a multi-gas sensing system.Withthe differential absorption method,the methane and carbon dioxide detection system was designed,constructed and tested.In the experiments,the methane detection sensitivityreached 271 ppm.The experimentalresults of carbon dioxide detection wereobservedwith high concentrations.The performanceof multi-point channels in the network was further evaluated.To resolve the issues and problemsin the above-mentioned optical fiber gas sensing architecture,this workproposes and demonstratesanotherrkind of multi-gas long distance sensing network based on photoelectric modulation.It combines the advantages of mature electric sensors with the superiority of long haul optical fiber communication network.A feedback neural network algorithm based on error(BP neural network)software platform was establishedto eliminate the cross interference.The sensing network spansmore than 25 km distance and is with multiple channels.In the experiment,three kinds of mixed gas,i.e.methane,carbon dioxideand hydrogen,was successfully detected,alongwith the cross interference cancelling technique.The measurement sensitivity approaches about 100 ppm.Considering the temperature variation influence,the concentration of system error further reduced to about 4ppm.Hence,the network performance is favourable with the overall fluctuation around 1% with the 100 ppm sensitivity.The reaearch workprovides an important and promising route for multi-gas detection network architecture.