| The effective exploitation of coal has provided a great deal of disposable energy for our country,but mine gas explosion accidents occurred frequently in the mining process,which has badly damaged the property security of the country and threatened the life as well as health of the miners.The main components of mine gas are CH4,H2 and CO,and their existence is a potential danger of mine gas explosion.Therefore,real-time accurate measurement is a crucial step to prevent mine gas explosion,which has the great practical significance and long-term social-economic benefits for achieving safety production of coal mine and protecting personal safety of miners.For the development of a mine gas quantitative detection system with high precision using MoS2 based sensor array combining with intelligent recognition algorithm,this thesis performed the following research work.Firstly,metal oxide(MOx)and MoS2 were synthetized by hydrothermal method,which are used as sensing materials towards mine gas.The MoS2 based gas sensors modified with MOx and noble metal were fabricated via two-step hydrothermal route and self-assembly technique.The surface morphology,microstructure,pore size as well as other characteristics were fully analyzed by using XRD,SEM,TEM and other characterization techniques.Secondly,the gas sensing properties of the Pd-SnO2/MoS2 sensor towards H2,the Ag-ZnO/MoS2 sensor towards CO and the Pt-Co3O4/MoS2 sensor towards CH4,were investigated from various aspects,including dynamic switching response,response/recovery time,repeatability,stability as well as selectivity.Based on band structure of heterojunction,charge transfer of thin film,physical-chemical properties of sensing material and catalytic performance of noble metal,the possible sensing mechanism for ternary composite film response to corresponding target gas was discussed in detail.Furthermore,the MoS2 based gas sensor array was constructed by high-performance sensor devices,and its cross sensitivity response towards the multi-component mine gas was obtained and analyzed.Finally,under the optimization of genetic algorithm(GA),three models of GA-BP,GA-RBF,GA-LSSVM were established and applied to the quantitative detection of mine gas.From the comparison results of model performances,GA-LSSVM model is superior to the other two ones and exhibits a good prospect in the rapid and high-precision monitoring of mine gas.Based on the research route of "new material-new device-new technology-new application",this paper focused on the quantitative detection of multi-component mine gas,which provides a certain theoretical guidance and practical reference for the development of high-preformace mine gas on-line sensing devices.Meanwhile,this work has important significance in preventing mine gas explosion and promoting the safe operation of the mining industry. |
PDF Full Text Request