On the calculation of forecasting parameters of gas-dynamic phenomena from spectra of seismo-acoustic signals

Purpose. Study of possible shortcomings of the procedure for calculating the normatively accepted prognostic parameters of the current forecast of gas-dynamic phenomena in mines based on spectral characteristics of seismoacoustic signals caused by the action of rock-destroying mechanisms on a coalface. The methods. The studies that served as a basis for establishing the existing procedure for calculating prognostic parameters were analyzed. Computer modeling of spectral curves and calculations based on the spectra of prognostic parameters was performed. An assessment of the correspondence between the values of parameters and the type of spectral curves was made. Findings. The relationship between forecasting methods using analog equipment and procedures for obtaining numerical values of forecast parameters is demonstrated. The standard procedure for calculating forecast parameters, implemented in most outburst-hazardous mines in Ukraine, is examined in detail. It is found that this procedure, at least in some cases, cannot reflect the redistribution of the spectrum and can lead to contradictory values of the sought parameters. The possibility of using quantiles of the cumulative function of the spectrum to form forecast parameters was investigated. It is shown that quantile changes adequately reflect shape changes and spectra shifts. The originality. The presence of unstable spectra of seismoacoustic signals relative to the values of the accepted prognostic parameters has been proven. It has been established that the prognostic parameters calculated by the accepted procedure may not reflect real changes in the signal spectrum and, accordingly, be erroneous. It is proposed to build a forecast of a dangerous state based on the spectrum quantiles of seismoacoustic signals. Practical implementation. The study results can be used to correct the algorithm for calculating prognostic parameters according to the accepted procedure and to improve the forecast methodology by using spectra quantile estimates.