Bridge damage identification method based on multiple cross validation

Abstract

The invention discloses a bridge damage identification method based on multiple cross validation. The method comprises the following steps that: continuously arranging n pieces of distributed long-gauge-length fiber grating sensors along a bridge extension direction below a bridge; connecting all sensors to demodulation equipment to collect the long-gauge-length strain response of the bridge whena vehicle passes; carrying out permutation and combination on all sensors, taking any three sensors as a group, and carrying out intra-class second order difference on the correspondingly obtained long-gauge-length strain; solving an area surrounded by each group of strain difference curves and a time horizontal axis; carrying out statistics on the number of the sensor related to an exceptional damage index S* out of a threshold value range; and substituting the number into a damage calculation formula to obtain a structure damage degree [Beta]. By use of the method, through the multiple crossvalidation, the exception and the randomness of an individual sensor are avoided, an obtained result has a high confidence coefficient, and meanwhile, the sensor is higher in durability and accuracy.The method is suitable for the bridges of a simply supported beam and continuous beam structure, low cost and large range application can be realized, the potential damage of the bridge is found in time, and accidents are avoided.

Description

technical field [0001] The invention relates to a bridge damage identification method, in particular to a bridge damage identification method based on multiple cross validation. Background technique [0002] Due to the long-term reciprocating upper vehicle load, the bridge performance will gradually decline. In order to avoid potential accidents and ensure the safety of the upper passers-by and vehicles, it is necessary to carry out real-time maintenance of the bridge. At present, the commonly used method is to complete by manual inspection. Practice has found that manual inspection is inefficient, takes a long time, and is highly subjective. It can only find structural surface damage and cannot effectively distinguish internal damage. Therefore, the damage identification algorithm based on bridge health monitoring is particularly necessary. [0003] At present, many scholars have developed many damage identification algorithms, but a large part of them are based on structu...

AI Technical Summary

Problems solved by technology

In practical engineering applications, it is found that these methods are not sensitive to local damage of the structure, and it is difficult to detect damage

In order to solve such problems, a damage identification method based on time-domain information was proposed, but due to the interference of the external environment, the method often has problems such as false positives and missed negatives, and the confidence of the identification results is limited.

Images