Abstract
When conducting experimental studies, numerical values of quantities are often recorded, which are often measured with some error associated not only with the measurement technique, the error measurement of instruments, but also with the presence of noise. In this regard, the natural desire of researchers to minimize measurement errors and noise is understandable. To solve this problem, when processing experimental data, it is necessary to use the approximation of the results, i.e. smoothing the curve data. This article provides a comparative analysis of some of the most effective smoothing algorithms based on data obtained experimentally using a laboratory benchmark with removable sensors MQ-135 and ME2-O2-F20. The application of smoothing algorithms was implemented using the software application SCA (Smoothing Curve Application) which allowed to cancel hardware measurement noise.