Research On Compressed Sensing Technology Of Gas Data In Underground Coal Mine

In the process of coal mining,gas continuously emerges from coal seams or rock layers.Gas monitoring equipment can capture a large amount of gas data.If the gas data cannot be analyzed in time,coal and gas outburst and gas explosion will occur.Various hazards have seriously threatened the personal safety of underground personnel.Therefore,we urgently need to analyze the collected gas data in a timely and accurate manner to reduce the series of hazards caused by gas.How to compress and sense a large amount of real-time gas data in a limited storage space and to recover compressed gas data with high precision is crucial.In view of the low reconstruction accuracy and complex reconstruction process of gas data signals,this paper focuses on the signal sparsity and observation matrix construction techniques in gas signal compression sensing.The variation of the gas signal is more sparse and structured by variational mode decomposition.An adaptive observation matrix suitable for sparse signals greatly increases the reconstruction probability of gas data.The main research contents of this paper are as follows:(1)In the process of collecting gas data,a small amount of missing and abnormal data will be obtained,which directly affects the analysis and processing of the signal,and the missing and abnormal gas data is filled by the simple moving average method to obtain completeness.Gas data signal;since the gas data signal is large and non-sparse,the gas signal is separated by using Variational Mode Decomposition(VMD)to obtain a series of morphological function signal components of the gas signal.By setting the threshold to retain valid information and removing redundant information,the gas signal is more sparse.At this time,the gas data signal is sparse and contains most of the information of the original gas data signal.(2)According to the non-stationary and abrupt characteristics of the gas data signal,an adaptive observation matrix suitable for sparse signals is constructed.The Gaussian matrix is firstly block-processed,and then the threshold matrix and the selected sparse basis are selected.Do inner product processing to construct an adaptive observation matrix.The observation matrix can be changed with the sparse signal,and while retaining a large amount of information,the signal without the information amount is eliminated and the clear operation is performed.The comparison experiments show that the proposed algorithm has higher reconstruction quality.