The Identification Of Multi Component Gas Of Coal Spontaneous Combustion Based On TDLAS

The fire accidents caused by spontaneous combustion of coal are very serious in China.The relevant data indicate that the majority of coal fires originate from coal spontaneous combustion.Some easily detected,representative and regular index gases will appear when coal spontaneous combustion occurs,such as CO and CO2.By accurately detecting gas concentration,it can provide reliable data for predicting natural fire of coal and reduce accident losses.This paper uses tunable diode laser absorption spectroscopy(TDLAS)to achieve high precision and non-contact detection of gas.First,through introduction of theoretical analysis of the multi component detection of TDLAS coal spontaneous combustion starts the article.Then,the main components of the system are introduced,the overall structure of the gas monitoring system is designed,and experimental platform is set up to collect the concentration of five kinds of gas in CO,CO2,CH4,C2H4 and C2H2.Because CO is the most key gas to predict the degree of spontaneous combustion of coal,using formula correction method to do initial temperature and pressure compensation.Then,the CO gas concentration correction model of ELM(extreme learning machine),GA(genetic algorithm)optimization BP and PSO(particle swarm optimization)optimization BP are established respectively.By comparing the test set accuracy of the three modified models and the time available,the CO concentration correction model based on PSO-BP is stable and effective.The correction effect is the best and the time is shorter.It can be applied to the monitoring of gas concentration.Finally,the on-line monitoring system of CO gas concentration based on TDLAS is designed by using virtual software Lab VIEW.The system includes the data acquisition module,the gas concentration detection module based on the PSO-BP,and historical data query module.The online monitoring system of gas concentration designed in this paper provides theoretical reference and design guidance for the CO gas concentration real-time monitoring,the daily maintenance and management which has far-reaching application and popularization value.