The usage of statistical analysis methods for controlling the operational stability of gas treatment facility

The paper considers the solutions to the scientific and practical task of improving the accuracy and reliability of natural gas consumption by conducting a statistical analysis of the results of gas pressure measurements in the pipeline, obtained from three pressure sensors during 12 hours. The importance of this task is underlined by the fact that the confidence interval for RMS errors is usually very wide. Testing of the hypothesis of the instability absence in the process of measuring gas pressure is carried out using a single factor dispersion analysis (the equation of the median values), a linear regression analysis (no influence of time on the value of the indicator for each voter) and a covariance analysis (no difference in the functional influence of the time on the number of an indicator). Three series of the measurement results of X control indicator (gas pressure in the pipeline) have been used. The analysis has confirmed the hypothesis that there are no disruptions to the stability of the gas pressure measurement process, which makes the pressure sensors metrologically reliable. It has been proved that the scientific and applied problems of increasing the reliability of objects control and diagnostics with stochastic parameters and improving their metrological reliability are relevant and important for the development of the theory and practice of non-destructive control and functional diagnostics of objects. Keywords: gas equipment; metrological reliability; statistical evaluation.