Application Research Of Two-Stage Data Fusion Algorithm In Coal Mine Dust Monitoring

With the deepening of the coal mine production automation,the problems of large quantity,complicated ingredients and the difficulty in management about coal mine dust had been presented.As an advanced dust reduction device,the coal dust online monitoring system had realized the real-time monitoring of dust concentration.However,it still has many wrong decisions and false actions because of the single monitoring indicator and a large number of redundant error messages,all these lead to the low efficiency of dust removal and waste of water resource.Therefore,the two-stage data fusion algorithm was applied to the coal dust monitoring system in this paper.In order to eliminate the redundant information,the original data of homogeneous sensors was filtered out by using the data layer(first stage)fusion algorithm.Becase of algorithm of single and insufficient application of D-S evidence theory and RS theory,a new decision layer algorithm based on D-S evidence theory and RS theory was proposed.The feasibility of the algorithm was proved by theoretical analysis.Moreover,thefield applications in different industrial and mining locations were also carried out.The main research contents of this paper are summarized as follows:(1)In order to make a comprehensive and effective dust removal decision,based on the analysis of selection of sensor monitoring indicators and data process,the monitoring indicators were divided into total dust concentration,respirable dust concentration,dust particle size and wind speed.The coal dust monitoring database was established,which layered the foundation for the application of two-stage data fusion algorithm in coal dust monitoring system.(2)In order to adapt to the complicated environment and changeable working condition,the two-stage structure model of data fusion algorithm was established.The algorithm which based on support matrix was used to data layer fusion.At the same time,the algorithm idea was clarified,algorithm steps were presented and the example of dust monitoring database was given.Based on the analysis of the algorithm based on D-S evidence theory and the algorithm based on RS theory,the decision results of the two algorithms were compared through the example of coal dust monitoring.Because of problems of single decision rule and some other reasons,it was found that the decision result was not accurate enough and the reliability was not high.(3)In this paper,a new decision layer algorithm based on D-S evidence theory and RS theory was proposed,and the feasibility of the algorithm was proved by theoretical analysis.In order to further improve the theoretical basis and operational rules of the new algorithm,three key problems(BPA functionacquisition,attribute importance assessment,evidence synthesis method)in the process of the new algorithm were analyzed.Finally,it was proved that the new algorithm is more advantageous than other two algorithms.It can not only objectively determine the BPA function and output the decision results in the form of probability,but also improve the accuracy and reliability of dust decision obviously.(4)Firstly,the application of two-stage data fusion algorithm in coal dust monitoring was verified by examples,then the effect of two-stage data fusion algorithm on the coal dust monitoring system was studied based on the dust control and water saving effect in the test coal mine.Results showed that under the premise of ensuring the effect of dust control,the working time of the improved system spray device is reduced by about 30% compared with the unimproved one,and the spray volume is decreased by about 20%.Field application results showed that the application of the data fusion algorithms has improved the function and efficiency of the dust on-line monitoring system,which proved that the proposed method is correct and feasible.