Spectral Analysis of Signals with Uneven Sampling Using the Example of Preprocessing of Profilograms

Spectral analysis of experimental signals measured with uneven sampling gives an error that does not allow practical use of the results. An algorithm and a computer program for converting a signal with an uneven measurement pitch into a signal, with a uniform pitch (“pitch equalization”), have been developed. The signals are represented by coordinate series. The algorithm is based on the symbolic Dirac unit impulse function of a real variable and its approximation by a continuously differentiable function. The algorithm was tested on the experimental results of measuring the profilogram of the outer surface of a cylindrical sample processed by turning. The results of spectral analysis of profilograms with a uniform measurement step are given.