A Bolt Preload Identification Method Based on Singular Value Decomposition and Support Vector Machine

Abstract

The invention provides a bolt pre-tightening force identification method based on singular value decomposition and a support vector machine, and belongs to the technical field of vibration signal fault detection. The invention provides a bolt pre-tightening force measuring method based on a vibration signal in order to solve the problems that the pre-tightening force of a bolt is reduced and eventhe bolt is separated and falls off due to the fact that mechanical environments such as impact and vibration generated in the operation process of mechanical equipment. The method comprises the stepsof collecting an impact vibration signal of a single bolt connecting beam in a free state through DASP modal analysis software; carrying out empirical mode decomposition and reconstruction on the experimental data to remove noise and trend terms; carrying out feature dimension reduction by combining singular value decomposition, and extracting a vibration signal feature vector; and taking the vibration signal feature vector as an input feature, performing classifier model training through the support vector machine, and finally realizing pre-tightening force identification of the bolt connecting structure by using the trained classifier.

Description

technical field [0001] The invention relates to a bolt pretightening force identification method based on singular value decomposition and support vector machine, and belongs to the technical field of vibration signal fault detection. Background technique [0002] Bolted connection structures are widely used in various engineering structures, and their connection status directly affects the safety and reliability of the entire structure. The pre-tightening force is one of the most important parameters for the bolted connection structure, and how to accurately control the pre-tightening force of the bolted connection structure is an urgent problem to be solved, both in the theoretical research process and in the practical engineering application. Precise control of preload force can be generally divided into two steps. First, accurate identification of bolt preload force is required, followed by accurate control of preload force. [0003] The existing bolt pretightening forc...

AI Technical Summary

Problems solved by technology

[0005] In view of the problems referred to above, the object of the present invention is to provide a kind of preload identification method that combines the singular value decomposition feature extraction method and support vector machine for the vibration signal of the connecting structure, that is, based on the vibration signal singular value decomposition and support vector machine Bolt pre-tightening force identification method to solve the problems of existing bolt pre-tightening force identification methods in practical engineering applications such as high processing cost, easy to be affected by the environment, complex measurement process, and weak adaptability

Images