Abstract
<jats:p>The article presents the results of the research work on testing CRIN2 electromagnetically controlled fuel injectors of various operating times using a specialized diagnostic bench and a multichannel measuring system with a flexible structure. The original signal was decomposed and the oscillation amplitudes were recorded, the characteristic oscillation frequencies of the injector were determined, on the critical parts of which traces of wear were detected. In total, the array of spectrograms and wavelet transforms of vibration signals of CRIN2 generation injectors obtained using the MATLAB mathematical package made it possible to form and fill the database. It is advisable to use this database for in-depth analysis of spectrograms and wavelet transforms using neural networks with the implementation of machine learning of the descriptor.</jats:p>