Multi-scale Spectrum Of On-line Optical Monitoring And Intelligent Identification Of Pollution Source In Driving Working Face

The 13 th Five-Year Plan for the development of coal industry calls for the establishment and perfection of long-term mechanism for coal mine safety production and the improvement of workers’ health status.However,the current situation of coal mine safety production is still very serious in China.Because of the blasting operation in mine heading face,large quantities of toxic and harmful gases and mine dust such as carbon oxides,nitrogen oxides,will be produced instantaneously.They will spread to the whole tunneling and mine ventilation network with ventilation which pose a major threat to the safety of mine production and the safety of workers.Therefore,it is of great practical significance and research value for ensuring the long-term safety of tunneling operation and the occupational health of workers,and realizing intelligent ventilation and emergency control in modern mines to build an on-line monitoring system of mine heading face,master the dynamic diffusion law of flue gas hazard sources,perceive and provide risk prevention and control references in real time.In view of this,this paper takes online optical monitoring and intelligent identification of pollution sources in 12182 heading face of Shihao Coal Mine as the research subject.Firstly,according to the current situation of the working face,the shortcomings and practical needs of the on-site supervision are analyzed.Based on the ultraviolet-visible and infrared absorption gas concentration measurement method and the bundle tube monitoring technology,a distributed on-line optical monitoring system for the driving face is designed.Secondly,in order to meet the needs of adjusting and evaluating the online monitoring system of pollutant concentration,the numerical method is used to reproduce the characteristics of smoke diffusion process and transient distribution in heading face,meanwhile,the large simulation data of the spatio-temporal evolution characteristics of pollutant sources are established.Finally,by studying the multi-scale spectra of harmful gases(ultraviolet-visible nano-scale,infrared micron-scale)and combining the large data of pollutant distribution,a mapping model between multi-component concentration and light radiation characteristics is set up.In addition,the time-domain characteristics of optical monitoring signals of pollution sources and their influencing factors are analyzed.More importantly,artificial intelligence algorithm is introduced to realize high precision non-linear correction of sensor and real-time reconstruction of pollution source concentration distribution.Through the above research,a distributed dual-band collaborative online monitoring system and an optical intelligent perception and identification method of pollution sources for mine heading face are proposed in this paper.The research findings are able to provide design bases and technical supports for automation and intelligent upgrade of modern mine monitoring.