Sparse wavelet decomposition in problems of vibration-based diagnostics of rotary equipment

Evaluation of the technical condition of rotary equipment on vibration signals is traditionally important task in the practice of its production and operation. At the present an increase in the effectiveness of vibration-based diagnostics is achieved by automating the solution of this problem and also by the use of matched sets of informative features, which causes the urgency of the development of algorithms for their detection. The algorithm of sparse wavelet decomposition is proposed to detect shock processes in vibration signals. The algorithm is tested on the experimental material. The results of the experiment are compared with the matched wavelet filtration.