Plate structure impact monitoring method based on fractal theory

Abstract

The invention discloses a plate structure impact monitoring method based on the fractal theory, and belongs to the technical field of impact monitoring of structure health monitoring. The method comprises steps as follows: step one, distribution of distributed fiber bragg grating sensors; step two, computation of box-counting fractal dimensions of the fiber bragg grating sensor impact response signals; step three, collection of the fiber bragg grating sensor impact response signals corresponding to sample impact points and box-counting fractal dimension computation; step four, collection of the fiber bragg grating sensor impact response signals corresponding to to-be-detected impact points and box-counting fractal dimension computation; step five, computation of x-axis coordinates of the to-be-detected impact points; step six, computation of y-axis coordinates of the to-be-detected impact points. According to a positioning algorithm, the positions of the impact points are judged through computation of the box-counting fractal dimensions of fiber bragg grating sensor impact response signal frequency spectrums, and the method has the characteristics of no need of a large number of prior knowledge, simplicity, rapidness, convenience, high practicability and the like.

Description

technical field [0001] The invention belongs to the technical field of impact monitoring for structural health monitoring, in particular to a plate structure impact monitoring method based on fractal theory. Background technique [0002] Modern large-scale structures such as space shuttles, high-rise structures, new bridges, and long-span grid structures will be affected by design loads and various sudden external factors such as foreign object impact and vibration in complex service environments. These vibrations or shocks can cause varying degrees of damage to the structure. These damages are highly concealed, the mechanism of damage and failure is complicated, and the type and degree of damage are difficult to judge. However, if these damages are not discovered in time and corresponding measures are not taken, they will cause structural damage accumulation, degrade the mechanical properties of the structure, and threaten the safety of the structure. , with disastrous con...

AI Technical Summary

Problems solved by technology

These damages are highly concealed, the mechanism of damage and failure is complicated, and the type and degree of damage are difficult to judge. However, if these damages are not discovered in time and corresponding measures are not taken, they will cause structural damage accumulation, degrade the mechanical properties of the structure, and threaten the safety of the structure. , with catastrophic consequences

[0003] The research on the positioning and identification of shocks mainly focuses on two directions: one is based on the Lamb wave excited by the shock in the plate structure, and the representative ones are triangular positioning, quadrilateral positioning and time difference positioning based on wavelet transform, etc. A good impact position identification result has been obtained, but due to the anisotropy of the plate structure, the propagation velocity of the Lamb wave propagating in it varies with the frequency and direction of propagation, which directly reduces the accuracy of positioning based on the Lamb wave

In addition, these methods are easily affected by electromagnetic interference due to their data acquisition and transmission process, and the piezoelectric monitoring array requires a large number of transmission cables, which greatly increases the complexity of the system. Representative ones are position recognition based on Hausdorff distance and position recognition based on support vector machine, etc., but these algorithms are greatly affected by their own characteristics. For example, the positioning accuracy of support vector machine algorithm depends on the selection of kernel width and penalty factor , but the choice of kernel width and penalty factor is not supported by relevant theories, and can only be determined by experimental trials or other methods. If the parameter selection is inappropriate, the positioning accuracy will be reduced

In addition, there is a neural network impact location identification method based on a large amount of prior knowledge. The disadvantage of this type of method is that it needs a large number of different impact samples to train the network, the generalization performance is not good, it is easily affected by temperature changes, and the workload is large. Its practicality and real-time performance are limited

Images