Resumen
The paper deals with the connection between the issue of rectification of the structural system and the corresponding experimental and related statistical analysis of stresses. The solution was demonstrated in the example of replacing the aboveground gas pipeline with an underground solution. During the implementation of landscaping for the underground solution, there was a risk of damage to the existing structure. Due to the specific problems in stress analysis based on strain gauge measurements, especially with regard to the narrow range of gas pressure as the primary input quantity, the obtained data had to be subjected to complex statistical procedures. Several linear and non-linear regression models were compiled and subjected to normality, homoscedasticity, and autocorrelation tests. The data obtained in this manner ruled out the possibility of limited state occurrence on the existing structure and, in conjunction with other information, enabled the successful completion of the rectification process to its current form. The scientific added value of the contribution should be the demonstration of procedure methodology and relatively unconventional use of regression and correlation analysis methods and means in the technical field for stress analysis in solving engineering practice problems.