A method for damage identification of bolted joints based on time series analysis and information entropy

Abstract

The invention discloses a bolting joint part damage recognition method integrating time series analysis and information entropy, and the method comprises the following steps: 1), arranging a plurality of acceleration sensors on a bolting joint part structure, so as to achieve the actual measurement of the acceleration response sequence of corresponding measurement points; 2), calculating the standard deviation of the acceleration response sequence of all measurement points on the structure, taking the maximum value of the standard deviation as a normalization parameter, carrying out the normalization of the actual measurement acceleration response, and selecting a proper time series model from an AR model, an ARMA model and an MA model for data fitting according to the characteristics of the actual measurement acceleration response; 3), calculating the residual error sequence between the actual measurement acceleration response data and a fitting model, and taking the standard deviation of the residual error sequence as the structural response features; 4), introducing information entropy to quantifying the relation between the standard deviations of the residual error sequences of two adjacent measurement points, calculating the structural damage sensitive index in a to-be-measured state, and recognizing the damages to the bolting joint part structure. The method can achieve the accurate recognition of the damage state of the joint part, and meets the requirements for the monitoring of the connecting state of the bolting joint part.

Description

technical field [0001] The invention relates to the field of structural health monitoring, in particular to a damage identification method for a bolted joint combining timing analysis and information entropy. Background technique [0002] For the health monitoring system of engineering structures, the monitoring of the connection state of bolted joints is very important. It is a key component for transferring loads between connectors and affects the operational safety of structures. However, due to the complex mechanism of the bolted joint, it is very difficult to describe the real mechanical behavior of the joint through nonlinear theory or linear theory, and the modeling is expensive. In this context, it is very important to study the model-free method for damage identification of bolted joints. The mentioned model-free method refers to: no need to establish a mechanical model of the joint, the structural response signal is obtained through the sensor installed on the str...

AI Technical Summary

Problems solved by technology

The idea of ​​constructing the existing indicators is usually to examine the degree of deviation of the structure’s state response from the reference state, but no matter the damage of the connector or the damage of the joint will cause the structural state to deviate from the reference state, so it is easy to misjudge the damage of the joint

Images