Research And Implementation Of A Multi-source Sensing-based Early Warning Mechanism For Coal Mine Gas Monitoring

Coal occupies the main position in China’s primary energy production and consumption,and is an important support for the development of our national economy.The safe production of coal has always been a topic of concern.In recent years,with the rapid development of economic and computer technology,in the background of information technology,coal mine information technology and automated production have also been attached importance by countries and various enterprises.The proportion of coal mine safety accidents caused by coal mine gas has been high,and although the trend of change in the death rate of one million tonnes in China’s coal mine production safety accidents has decreased in recent years,there is still a gap compared to developed countries in Europe and the United States.Therefore,an accurate coal mine gas early warning mechanism is the foundation and prerequisite for safe coal mine production.In this paper,based on the data collected by multi-source sensing,the Sparrow Search Algorithm(SSA)is improved and fused with Support Vector Machine(SVM)to achieve coal mine gas safety warning,and a BP neural network based adaptive localization algorithm is proposed to improve the accuracy of the localization algorithm.The specific research work and results are as follows:(1)To address the problems of uneven distribution and the tendency to fall into local optimality during the search process of the sparrow search algorithm in initializing the population,this paper solves the problem of uneven initializing the population by introducing Bernoulli mapping and reverse learning,introduces the Corsi variation to jump out of the local optimal solution,and obtains the global optimal solution through the simplex method to obtain the improved algorithm BCMSSA,and establishes the BCMSSA-SVM coal mine gas safety The improved algorithm BCMSSA is obtained and the BCMSSASVM coal mine gas safety warning model is established.In this paper,the performance of the improved sparrow search algorithm is examined through seven sets of comparative experiments.Compared with other intelligent optimisation algorithms and improved algorithms,BCMSSA has better convergence performance and optimisation finding ability;four sets of experiments are designed to verify the effectiveness of the BCMSSA-SVM coal mine gas safety warning model.(2)For the problem of inaccurate positioning of people in special environments,a BP neural network based adaptive positioning algorithm is proposed.The algorithm firstly pre-processes the multi-source data and improves on the traditional localisation algorithm to initially improve the localisation accuracy.Secondly,the influence of the environment on the signal is taken into account in the localisation technique in special scenarios,and the environmental adaptive path loss coefficient is obtained through the BP neural network with the environmental parameters as the input conditions to further improve the accuracy of the localisation algorithm.Through six sets of controlled experiments.the performance of the BP neural network-based adaptive localisation algorithm is verified and the accuracy of localisation is effectively improved.(3)Based on the BCMSSA-SVM model and BP neural network adaptive positioning algorithm,this paper designs and develops a B/S architecture based coal mine gas monitoring and early warning system.The software is developed using Spring Boot and Vue technology,and the MQTT protocol is used to communicate the data.The coal mine gas monitoring and warning system meets the requirements of gas safety warning and personnel location.