Mechanical vibration signal data compression and acquisition method

Abstract

The invention, which belongs to the technical field of the mechanical vibration signal data acquisition method based on the compression sampling theory, discloses a mechanical vibration signal data compression and acquisition method. With a measurement matrix, a mechanical vibration signal is projected into a low-dimensional compression collection value, a sparse optimization problem is solved at a vibration signal reconstruction terminal, and a high-dimensional mechanical vibration signal is reconstructed by using the low-dimensional compression collection value with high probability. According to the invention, a problem of mass mechanical vibration data obtaining based on a traditional sampling theorem being a Shannon-Nyquist method can be solved. The measurement matrix carries out compression sampling on the mechanical vibration signal after analog-to-digital conversion and the compression sampling value is transmitted, stored and processed, so that the original mechanical vibration data volume is substantially reduced. Therefore, a high signal compression ratio can be obtained and the precise signal reconstruction performance can be realized.

Description

technical field [0001] The invention relates to a mechanical vibration signal data compression acquisition method, which belongs to the field of vibration signal data acquisition and compression methods, and specifically belongs to the technical field of a mechanical vibration signal data acquisition method based on compression sampling theory. Background technique [0002] Mechanical vibration signals transmit and carry a large amount of important information contained in the working process of mechanical equipment. On-line monitoring and collection of mechanical vibration signals is one of the key technologies in the field of mechanical engineering, especially in fault diagnosis technology. The traditional and currently widely used sensors for vibration signal detection and sampling have always been based on the Shannon-Nyquist sampling theorem, such as figure 1 As shown, the classic Shannon-Nyquist vibration signal acquisition theory requires that the acquisition frequenc...

AI Technical Summary

Problems solved by technology

However, with the continuous improvement of actual industrial production requirements and the continuous development of the machinery industry, mechanical equipment is becoming more and more large-scale, complete sets, high-speed, networked, integrated, etc.; Sudden changes, structural deformation, friction changes, etc., and cross-coupling between different components, the vibration is more complex, the random vibration frequency is getting higher and higher, and it presents nonlinear and non-stationary

Moreover, with the development of large-scale mechanical vibration monitoring towards comprehensive, high-speed, continuous and networked trends, vibration signal sampling based on Shannon-Nyquist has resulted in a huge amount of sampled data, and the real-time transmission and synchronous storage of these data has become an urgent need. Solve the bottleneck problem of cost and engineering technology, especially in the field of Internet remote equipment status monitoring, a large amount of collected data has caused huge pressure on transmission

[0003] There are two defects in the existing collection method: first, the way of sampling first and then compressing is a serious waste of sampling resources; second, if the transformed and compressed data is lost during transmission and storage, the original vibration signal will be damaged. unrecoverable problem

Images