Research On Anti-aliasing And Real-time Early Warning Method Of Coal Mine Disaster Drift Characteristics

Coal resources play an important role in industrial production.In recent years,the demand for coal resources in various industries is increasing,the depth and intensity of coal mining is also increasing,and the frequency of rock burst appears more and more.When the impact ground pressure occurs,the underground roadway or the mining face can be destroyed instantly,and the high gas mine has the impact and low pressure,and it will also cause secondary disasters such as gas outburst,gas explosion and so on.Therefore,how to detect the impact ground pressure safely and effectively has always been a hot spot of experts in the field of coal mine microseismic monitoring technology.At present,most of the collection and storage of mine disaster data are carried out by using sensor base stations set around the mine.The microseismic signals of coal mines play an important role in the study of vibration characteristics,attenuation law and disaster assessment.However,due to the influence of geological structure,energy loss and other factors,the propagation of microseismic signal of the same kind of disaster in time domain will appear data drift phenomenon such as slowing or aggravating,which leads to the stretching or compressing of the perceptual disaster waveform.Furthermore,the accuracy of early warning of shock ground pressure is affected.Therefore,this paper studies the anti-aliasing and real-time early warning methods of coal mine disaster characteristics.First of all,aiming at the problem of data drift in the process of transmission of disaster sensing data,the anti-aliasing processing method of coal mine disaster drift characteristics is proposed.By introducing dynamic time warping algorithm(DTW)in the field of audio frequency recognition,anti-aliasing processing of real-time microseismic sensing data is carried out,and similarity fitting between real-time microseismic sensing data and historical disaster data template waveform is carried out to identify whether it is a disaster signal or not.According to the uncertainty of coal mine disaster,in the process of judging whether the perceptual data is a disaster signal and fitting the similarity,in order to determine the starting point of the matching between the perceptual data and the disaster template waveform,the real-time matching is realized.In this paper,a variable sliding window strategy is proposed to achieve waveform alignment.The initial alignment position of the perceptual disaster wave and the template wave is found by the gradual sliding of the perceptual window,and the window size is constantly adjusted to determine the disaster duration.Then ensure the accuracy of signal contrast.Secondly,because of the sudden problem of coal mine disaster,this paper puts forward the real time early warning mechanism of coal mine disaster based on the window matching strategy and the characteristics of mine disaster waveform to improve the real time of early warning.Firstly,the perceptual data is extracted by the method of piecewise cumulative approximation,and the computation is reduced.Then,based on the hierarchical early-warning mechanism,the contrasting waveforms are compared in sections,and the 1/N window size is used as the warning threshold to improve the warning level step by step to monitor disaster information in a short period of time.Finally,according to the method of coal mine disaster determination proposed in this paper,experimental analysis is carried out,which is verified by setting the sliding value and window variation value of different sizes respectively.By comparing the full window early warning with the 1/N window early warning,the efficiency of the proposed early warning method is verified.The experimental results show that the proposed early warning method has higher accuracy and real time performance.