CN-122017046-A - Defect detection method and system for precast reinforced concrete pile

CN122017046ACN 122017046 ACN122017046 ACN 122017046ACN-122017046-A

Abstract

The invention relates to the technical field of industrial automation monitoring and control, in particular to a defect detection method and system of a precast reinforced concrete pile, wherein the method comprises the steps of obtaining a pile body vibration signal and decomposing a wavelet packet; the method comprises the steps of calculating kurtosis values and energy concentration factors of all nodes, fusing the kurtosis values and the energy concentration factors to obtain defect likelihood values, performing self-adaptive optimal basis searching based on the defect likelihood values to determine optimal bases for highlighting defect signals, reconstructing the defect signals based on the optimal bases, and accordingly extracting and analyzing defects. The invention can effectively separate the defect characteristics from the strong background interference and trigger the response action through multi-characteristic fusion and self-adaptive search, thereby improving the self-adaptive capacity, detection sensitivity and quality monitoring automation level of the monitoring device in a complex industrial environment.

Inventors

GUO ZEKUN
GUO ZHITAO
WANG LIBING
LIU HONGLING
LI YUNTAO
FU YANCONG

Assignees

山东中能杆塔有限公司

Dates

Publication Date: 20260512
Application Date: 20260204

Claims (10)

1. The defect detection method of the precast reinforced concrete pile is characterized by comprising the following steps of: Obtaining a vibration response signal of the pile body, and carrying out wavelet packet decomposition on the vibration response signal to obtain a wavelet packet full tree comprising a plurality of nodes, wherein each node is a coefficient sequence; Calculating the kurtosis value of the coefficient sequence of each node; For each node, calculating an energy concentration factor of the coefficient sequence based on the energy entropy and the total energy of the coefficient sequence; Calculating a defect likelihood value of each node based on the kurtosis value and the energy concentration factor, wherein the defect likelihood value is the product of the square of a normalized value of the energy concentration factor and the logarithmic value of the kurtosis value when the kurtosis value is larger than zero, and is the product of the normalized value of the energy concentration factor and the kurtosis value when the kurtosis value is not larger than zero; and carrying out sparsification processing on the wavelet packet whole tree based on the optimal basis to obtain a sparse wavelet packet Quan Shu, carrying out inverse wavelet packet transformation on the sparse wavelet packet whole tree to obtain a defect signal, and generating a corresponding control instruction according to the defect signal.
2. The method for detecting defects of precast reinforced concrete piles according to claim 1, wherein the calculating an energy concentration factor of the coefficient sequence for each node based on the energy entropy and the total energy of the coefficient sequence comprises: Calculating the total energy of the coefficient sequence; Calculating the energy entropy of the coefficient sequence; the energy concentration factor is the product of the negative natural index of energy entropy and the total energy.
3. A method of defect detection for precast reinforced concrete piles according to claim 2, wherein said calculating the energy entropy of said coefficient sequence comprises: Calculating the energy ratio of the energy of each coefficient in the coefficient sequence to the total energy of the sequence; And calculating shannon entropy of the coefficient sequence based on the energy duty ratio to obtain the energy entropy.
4. The method for detecting defects of precast reinforced concrete piles according to claim 1, wherein the performing an adaptive optimal basis search on a wavelet packet full tree based on the defect likelihood values, determining an optimal basis, comprises: recursion decisions from the bottom layer of the wavelet packet full tree up layer by layer; Comparing the sum of the defect likelihood values of the father node and the defect likelihood values of two child nodes of the father node aiming at any father node; When the defect likelihood value of the father node is not less than the sum of the defect likelihood values of the two child nodes, reserving the father node and excluding the two child nodes; And when the defect likelihood value of the father node is smaller than the sum of the defect likelihood values of the two child nodes, the father node is eliminated and the two child nodes are reserved.
5. The method for detecting defects of precast reinforced concrete piles according to claim 1, wherein the obtaining of the defect likelihood values comprises: Taking any node as the current node, wherein the defect likelihood value satisfies the relation: ; Wherein, the Is the current node Is a defect likelihood value of (1); Is the current node Is a factor of energy concentration; Is the current node Kurtosis value of (a); Is the maximum value of the energy concentration factors in all nodes in the wavelet packet full tree; is a preset third minute value.
6. The method for detecting defects of precast reinforced concrete piles according to claim 1, wherein the thinning of the wavelet packet full tree based on the optimal basis comprises: and setting all node coefficients which do not belong to the optimal base in the wavelet packet full tree to zero, and reserving nodes in the optimal base.
7. The method for detecting defects of precast reinforced concrete piles according to claim 1, wherein the generating corresponding control instructions according to the defect signals comprises: When the pulse with the amplitude exceeding the preset threshold exists in the defect signal, judging that the pile body has defects, and calculating the depth of the defects according to the arrival time of the pulse and the sound velocity of the material of the pile body; triggering a control instruction based on the defect depth and amplitude.
8. The method for detecting defects of precast reinforced concrete piles according to claim 1, wherein the obtaining of the kurtosis value comprises: any node is taken as the current node, and the kurtosis value of the current node meets the relation: ; Wherein, the Is the current node Kurtosis value of (a); Is the current node Of the coefficient series of (2) An element; Is the current node The total length of the coefficient sequence of (a); 、 respectively the current node Average value and standard deviation of all coefficients in the coefficient sequence; Is a preset correction term.
9. A method of defect detection for precast reinforced concrete piles according to claim 1, wherein the acquiring the vibration response signal of the pile body comprises: the original vibration signal of clump of pile body is adopted; And filtering the original vibration signal to obtain the vibration response signal.
10. A system for defect detection of precast reinforced concrete piles, characterized in that it comprises a processor and a memory, said memory storing computer program instructions which, when executed by said processor, implement a method for defect detection of precast reinforced concrete piles according to any one of claims 1-9.

Description

Defect detection method and system for precast reinforced concrete pile Technical Field The invention relates to the technical field of industrial automatic monitoring and control, in particular to a defect detection method and system for a precast reinforced concrete pile. Background In the industrial automatic monitoring and nondestructive testing of precast reinforced concrete piles, the automatic acquisition device is used for transmitting stress waves to the pile body and analyzing echo signals received by the sensor to judge whether the pile body has defects such as cracks and hollows. However, in a complex industrial construction or production environment, the actually received echo signal is a highly complex non-stationary signal, which includes both a low-frequency global resonance signal of the pile body itself with strong energy and a high-frequency transient impact signal with weak energy caused by defects. Therefore, how to accurately extract weak defect features from strong background interference becomes a key point of monitoring system hardware and control logic design. In the prior art, a wavelet packet transformation equal-time frequency analysis method is generally adopted, complex signals are decomposed into a plurality of fine frequency bands for analysis, and the core is how to screen out optimal bases capable of highlighting defects from a large number of frequency bands. At present, the control logic of the monitoring device in the prior art generally depends on a single mathematical rule, and the logic threshold value is usually preset when leaving a factory and cannot be adjusted in real time according to the actual working condition. For example, the minimum shannon entropy criterion commonly used stems from information theory, which aims to find the most concentrated, information-ordered expression of signal energy. However, in the automatic monitoring detection scene of the precast reinforced concrete pile, the most concentrated energy part is usually the resonance signal of the pile body, and the transient impact caused by the defect is often misjudged as disordered noise under the stiff entropy rule logic due to extremely low energy ratio. Therefore, the unidirectional control logic based on the fixed parameters causes the monitoring system to lack self-adaptive adjustment capability aiming at complex working conditions, and finally the selected optimal base can submerge weak defect characteristics, so that the automatic monitoring device has low sensitivity and easy omission and judgment, and is difficult to meet the automatic requirement of the industrial field on-site full-flow quality control. Disclosure of Invention In order to solve the technical problems of weak self-adaptive control capability, stiff detection logic and difficult real-time adjustment caused defect missed judgment of the existing monitoring device in a complex industrial environment, the invention provides the following schemes in various aspects. In a first aspect, the present invention provides a method of defect detection of precast reinforced concrete piles, the method comprising the steps of: The method comprises the steps of obtaining a vibration response signal of a pile body, carrying out wavelet packet decomposition on the vibration response signal to obtain a wavelet packet full tree comprising a plurality of nodes, wherein each node is a coefficient sequence, calculating a kurtosis value of the coefficient sequence of each node, calculating an energy concentration factor of the coefficient sequence of each node based on the energy entropy and the total energy of the coefficient sequence, calculating a defect likelihood value of each node based on the kurtosis value and the energy concentration factor, when the kurtosis value is larger than zero, the defect likelihood value is a product of the square of a normalization value of the energy concentration factor and a logarithmic value of the kurtosis value, when the kurtosis value is not larger than zero, the defect likelihood value is a product of the normalization value of the energy concentration factor and the kurtosis value, carrying out self-adaptive optimal basis search on the wavelet packet full tree based on the defect likelihood value, determining an optimal basis, carrying out sparsification processing on the wavelet packet full tree based on the optimal basis to obtain a sparse wavelet packet Quan Shu, carrying out inverse wavelet packet transformation on the sparse wavelet packet full tree, and obtaining a corresponding control command signal according to the defect signal. The method comprises the steps of firstly carrying out preliminary screening on signals by using kurtosis values, distinguishing real defects and isolated noise in high kurtosis signals by using energy concentration factors, simultaneously carrying out normalization processing on the signals by introducing the maximum value of the energy concentration factors, co