Parallel steel wire cable broken wire identification method based on acoustic emission signals

Abstract

The application discloses a parallel steel wire cable broken wire identification method based on acoustic emission signals, relates to the technical field of bridge structure health monitoring and nondestructive testing, and comprises the following steps: a broken wire number characteristic fingerprint library is constructed; acoustic emission signals under different broken wire numbers are acquired through finite element simulation or experiments; and a characteristic fingerprint library containing energy ratio intervals and frequency band energy proportion interval corresponding to each broken wire number is established. The application effectively overcomes the signal interference problems caused by sensor position sensitivity and broken wire time difference through a frequency domain characteristic arbitration mechanism, and enhances the anti-interference capability. The method can complete monitoring only by using a single sensor, does not need to increase additional hardware costs, has outstanding engineering practicability and economy, is particularly suitable for long-term online monitoring of in-service bridges, and simultaneously, through the triple guarantee of characteristic matching, frequency band arbitration and scoring mechanism, it is ensured that a deterministic result can be output under various complex working conditions, fuzzy judgment is avoided, and the reliability of the system is greatly improved.

Description

Technical Field

[0001] This invention relates to the field of bridge structural health monitoring and non-destructive testing technology, and in particular to a method for identifying broken wires in parallel steel wire cables based on acoustic emission signals. Background Technology

[0002] Parallel wire cables are core load-bearing components of modern long-span bridges, and their safety is paramount. Acoustic emission technology, as a passive, real-time, dynamic, non-destructive monitoring method, has been widely used for the detection and location of cable breakage events. However, existing technologies mainly focus on determining "whether a breakage has occurred" and "where the breakage is located," while research on identifying "how many wires broke simultaneously" (i.e., the number of broken wires) in a single breakage event remains lacking.

[0003] Existing acoustic emission (AE) wire breakage detection technologies mainly rely on signal energy characteristics, waveform parameters, or time-frequency analysis. For example, the method and system for identifying bridge cable acoustic emission wire breakage signals under data imbalance (CN118395361A) proposes judging the occurrence of wire breakage events based on two-dimensional time-frequency images of the wire breakage signal; the literature "Acoustic Emission Monitoring of Main Cable Steel Wire Corrosion and Location of Wire Breakage Signal Source" uses wavelet packet decomposition to extract features of the wire breakage signal for wire breakage identification. However, these methods generally have limitations such as insufficient utilization of frequency domain features and a lack of identification mechanisms for "simultaneous wire breakage".

[0004] In practical engineering, multiple steel wires often break successively within a very short time (less than 0.1 ms) due to stress concentration or corrosion, forming a "simultaneous wire breakage event" with highly overlapping time-domain signals. Accurately identifying the number of broken wires is crucial for assessing the remaining load-bearing capacity and structural safety status of the cable. Existing technologies that directly use the energy characteristics of acoustic emission signals for wire count determination face two major technical challenges: first, the energy is sensitive to the sensor position, leading to misjudgments such as "the energy of two broken wires is lower than that of a single broken wire"; second, the signal interference effect caused by the small time difference between broken wires causes energy fluctuations, affecting the stability of the determination.

[0005] To address the aforementioned technical deficiencies, a solution is proposed. Summary of the Invention

[0006] The purpose of this invention is to solve the problem of recognition error caused by sensor position sensitivity and wire breakage time difference.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: a method for identifying broken wires in parallel steel wire cables based on acoustic emission signals, comprising the following steps:

[0008] Step 1: Construct a feature fingerprint database of broken wire count: Obtain acoustic emission signals under different broken wire counts through finite element simulation or experiments, and establish a feature fingerprint database containing the energy ratio range and frequency band energy proportion range corresponding to each broken wire count.

[0009] Step 2: Signal Acquisition and Feature Extraction: Deploy acoustic emission sensors and acquire acoustic emission signals from broken wires in parallel steel wire cables. Extract the total energy E of the signal and calculate the proportion of energy in the low-frequency band. Mid-frequency energy ratio and the proportion of high-frequency energy ;

[0010] Step 3, Feature Fingerprint Matching: Combine the energy ratio R=E / E1 and the proportion of low-frequency energy extracted in Step 2. Mid-frequency energy ratio High-frequency energy ratio Matching with the feature fingerprint database, where E1 is the energy reference value of a single broken wire;

[0011] Step 4, Frequency Band Arbitration Identification: When the matching in Step 3 fails, the number of broken wires is determined by arbitration rules based on the matching results of all frequency bands in the energy proportion of the low-frequency, mid-frequency, and high-frequency bands with the feature fingerprint database.

[0012] Step 5, Evaluation of the number of broken wires in the scoring mechanism: When the matching in step 4 fails, the energy and three frequency bands are scored. A conservative estimate is made based on the score, and if it is too high, the number of broken wires is determined.

[0013] Furthermore, in the feature fingerprint database, each number of broken wires corresponds to an energy ratio range and a low-frequency energy percentage range.

[0014] Furthermore, the feature fingerprint database specifically includes:

[0015] The number of broken wires indicates the number of broken wires in a parallel steel wire cable, and the value is 1, 2, 3, or 4.

[0016] The energy ratio range is used to represent the range of values ​​for the energy ratio of the acoustic emission signal with broken wires. It is divided into [0.8, 1.2], [2.0, 2.7], [3.1, 4.5] and [4.4, 6.8] according to the different numbers of broken wires.

[0017] The low-frequency energy percentage range is used to represent the range of values ​​for the low-frequency energy percentage of the broken wire acoustic emission signal. It is divided into [67.6%, 74.4%], [78.6%, 81.7%], [84.2%, 86.5%] and [89.1%, 91.2%] according to the different number of broken wires.

[0018] The intermediate frequency energy ratio range is used to represent the range of values ​​for the intermediate frequency energy ratio of the broken wire acoustic emission signal. It is divided into [24.7%, 31.3%], [17.7%, 20.8%], [13.1%, 15.3%] and [8.3%, 10.4%] according to the different number of broken wires.

[0019] The high-frequency energy percentage range is used to represent the range of values ​​for the high-frequency energy percentage of the broken wire acoustic emission signal. It is divided into [0.728%, 0.958%], [0.535%, 0.583%], [0.399%, 0.464%] and [0.355%, 0.364%] according to the different number of broken wires.

[0020] The characteristic fingerprint ID is used to name the different numbers of broken wires in parallel steel wire cables. The names are F1, F2, F3, and F4.

[0021] Furthermore, the low-frequency band ranges from 0 to 125 kHz, the mid-frequency band ranges from 156.25 to 250 kHz, and the high-frequency band ranges from 343.75 to 500 kHz.

[0022] Furthermore, the method for calculating the frequency band energy proportion is as follows: use the Db5 wavelet basis function to perform 5-level wavelet packet decomposition to calculate the energy proportion of each frequency band.

[0023] Furthermore, the process of arranging the acoustic emission sensor is as follows: the acoustic emission sensor is placed at the center of the uppermost cross-section of the parallel steel wire cable anchoring end.

[0024] Furthermore, the scoring mechanism is as follows: a total score of 100 points, of which energy accounts for 40 points, and high frequency, mid frequency, and low frequency each account for 20 points. The scoring mechanism specifically includes: energy ratio range, low-frequency energy proportion range, mid-frequency energy proportion range, high-frequency energy proportion range, corresponding root tendency, and score.

[0025] The corresponding number of wires is used to indicate the number of parallel wire cables with a strong tendency to break, with values ​​of 1, 2, 3, or 4 wires with a strong tendency.

[0026] The score is used to evaluate different numbers of broken wires. The score is calculated based on different energy ratio ranges and the corresponding wire number tendency. When it is in the transition range, the score is given proportionally.

[0027] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0028] This method for identifying broken wires in parallel steel wire cables based on acoustic emission signals achieves precise quantitative identification of the number of simultaneously broken wires by constructing a feature fingerprint database and introducing a triple discrimination mechanism. This elevates broken wire monitoring from the traditional "presence / absence judgment" to a new level of "quantity recognition," significantly improving identification accuracy. Secondly, the method effectively overcomes signal interference caused by sensor position sensitivity and broken wire time difference through a frequency domain feature arbitration mechanism, enhancing anti-interference capabilities. Furthermore, this method requires only a single sensor for monitoring, eliminating the need for additional hardware costs, demonstrating outstanding engineering practicality and economy, and is particularly suitable for long-term online monitoring of in-service bridges. Simultaneously, the triple guarantee of feature matching, frequency band arbitration, and scoring mechanisms ensures deterministic results under various complex operating conditions, avoiding fuzzy judgments and greatly improving system reliability. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0030] Figure 2 Box plot showing the total signal energy under the condition of 0-degree wire angle with the number of broken wires;

[0031] Figure 3 Box plot showing the total signal energy under the number of broken wires at a 30-degree wire angle;

[0032] Figure 4 Box plot showing the total signal energy under the number of broken wires at a 60-degree angle between the wires;

[0033] Figure 5 A graph showing the percentage of energy from the broken wire signal based on the number of single broken wires.

[0034] Figure 6 This is a graph showing the percentage of signal energy when two wires break simultaneously.

[0035] Figure 7 The diagram shows the energy distribution of the signal when three wires break simultaneously.

[0036] Figure 8 A graph showing the energy percentage of the signal when four wires break simultaneously.

[0037] Figure 9 This is a schematic diagram illustrating the impact of different wire breakage time differences on signal energy and time-frequency graphs. Detailed Implementation

[0038] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0039] Example:

[0040] like Figures 1-9 As shown, a method for identifying broken wires in parallel steel wire cables based on acoustic emission signals includes the following steps:

[0041] S1: Sensor optimization and signal acquisition. A single acoustic emission sensor is fixedly installed at the center of the uppermost cross-section of the parallel wire cable anchorage end using magnetic attraction or clamps. The preferred sensor is a broadband acoustic emission sensor with a resonant frequency of 150kHz and a frequency response range of 50-400kHz. This placement, based on findings from this invention, minimizes signal reception differences caused by random variations in the location of broken wires, as shown in the figure below regarding the placement of the acoustic emission sensor at the top of the cable anchorage end:

[0042]

[0043] During actual installation, petroleum jelly is applied between the sensor and the cable surface as a coupling agent to ensure good acoustic contact. The acoustic emission sensor is connected to the acoustic emission data acquisition device via a preamplifier (40dB gain), with the sampling rate set to 1MHz and the trigger threshold set to 40dB to avoid false triggering by ambient noise.

[0044] S2: Reference Signal Acquisition and Processing During the initialization phase of the monitoring system, at least 10 sets of acoustic emission signals from a single broken wire are acquired as reference data. Specifically, a single steel wire fracture is simulated in the laboratory or on-site, and the corresponding acoustic emission signals are acquired. The original signals are preprocessed using bandpass filtering (1kHz-500kHz) and wavelet denoising (Db5 wavelet basis) to calculate the reference energy value E1 for a single broken wire.

[0045] S3: Constructing a Feature Fingerprint Database Based on finite element simulation and experimental data, a feature fingerprint database of the number of broken wires is established as shown in the table below. The process of establishing this fingerprint database includes: obtaining acoustic emission signals under different numbers of broken wires (1-4 wires) through controlled experiments, statistically analyzing the distribution range of each feature parameter, and determining the feature intervals with discriminative power.

[0046]

[0047] S4: Real-time Signal Feature Extraction. When a wire breakage event is detected, the acoustic emission signal is acquired and the following feature parameters are extracted: total energy E; energy ratio R: R = E / E1; frequency band energy proportion: using the Db5 wavelet basis for 5-level wavelet packet decomposition, the signal is divided into 32 sub-bands, and the energy proportion of each frequency band is calculated, such as... Figures 5-8 As shown, where, Figure 5 (a) is a single broken wire. Figure 6 (b) consists of two broken wires; Figure 7 (c) consists of three broken wires. Figure 8 (d) represents four broken wires. The average energy of the broken wire signal for different numbers of broken wires is recorded in the table below:

[0048]

[0049] Simultaneously, the effects of different wire breakage time differences on signal energy and time-frequency diagrams were recorded, such as... Figure 9 As shown, where, Figure 9 (a) is a 0ms wire breakage time difference, (b) is a 0.1ms wire breakage time difference, (c) is a 0.2ms wire breakage time difference, (d) is a 0.3ms wire breakage time difference, and (e) is a 0.4ms wire breakage time difference.

[0050] S5: Feature fingerprint matching and discrimination: The feature parameters (R, ...) of the real-time signal are used to match and discriminate the fingerprint. , , It is matched and judged with the feature fingerprint database.

[0051] S6: Frequency Band Arbitration Mechanism: When fingerprint matching fails, the frequency band arbitration mechanism is activated to check... and Does it match the same range of roots? If so, then it is determined to be that root (highest priority).

[0052] S7: Scoring Mechanism When the arbitration mechanism still cannot determine the outcome, the scoring mechanism is activated. The specific scoring rules are as follows:

[0053] Total score 100, energy 40, high, medium and low frequency bands 20 each.

[0054]

[0055] Example of transition interval score (linear interpolation):

[0056] If R = 2.0 (between 1.8 and 2.8):

[0057] The score for 2 sticks = (2.0 - 1.8) / (2.8 - 1.8) × 40 = 8 points;

[0058] 1 point = 40 - 8 = 32 points;

[0059] 3 sticks and 4 sticks score 0;

[0060]

[0061] The transition period is also graded proportionally.

[0062] S8: Result Output and Storage. Outputs the number of broken wires identified, and records parameters such as timestamp, energy ratio, and frequency band percentage, generating a monitoring report. For identification results with low confidence, the system automatically labels them and suggests manual review.

[0063] This invention achieves precise quantitative identification of the number of simultaneously broken wires in parallel steel wire cables by constructing a feature fingerprint database and introducing a triple discrimination mechanism. This elevates wire breakage monitoring from the traditional "presence / absence judgment" to a new level of "quantity recognition," significantly improving identification accuracy. Secondly, the method effectively overcomes signal interference problems caused by sensor position sensitivity and wire breakage time differences through a frequency domain feature arbitration mechanism, enhancing anti-interference capabilities. Furthermore, this method requires only a single sensor for monitoring, eliminating the need for additional hardware costs, demonstrating outstanding engineering practicality and economy, and is particularly suitable for long-term online monitoring of in-service bridges. Simultaneously, the triple guarantee of feature matching, frequency band arbitration, and scoring mechanisms ensures deterministic results are output under various complex operating conditions, avoiding fuzzy judgments and greatly improving system reliability.

[0064] The size of the interval and threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by those skilled in the art for each set of sample data; as long as it does not affect the ratio between the parameter and the quantized value.

[0065] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0066] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for identifying broken wires in parallel steel wire cables based on acoustic emission signals, characterized in that, Includes the following steps: Step 1: Construct a feature fingerprint database of broken wire count: Obtain acoustic emission signals under different broken wire counts through finite element simulation or experiments, and establish a feature fingerprint database containing the energy ratio range and frequency band energy proportion range corresponding to each broken wire count. Step 2: Signal Acquisition and Feature Extraction: Deploy acoustic emission sensors and acquire acoustic emission signals from broken wires in parallel steel wire cables. Extract the total energy E of the signal and calculate the proportion of energy in the low-frequency band. Mid-frequency energy ratio and the proportion of high-frequency energy ; Step 3, Feature Fingerprint Matching: Combine the energy ratio R=E / E1 and the proportion of low-frequency energy extracted in Step 2. Mid-frequency energy ratio High-frequency energy ratio Matching with the feature fingerprint database, where E1 is the energy reference value of a single broken wire; Step 4, Frequency Band Arbitration Identification: When the matching in Step 3 fails, the number of broken wires is determined by arbitration rules based on the matching results of all frequency bands in the energy proportion of the low-frequency, mid-frequency, and high-frequency bands with the feature fingerprint database. Step 5, Evaluation of the number of broken wires in the scoring mechanism: When the matching in step 4 fails, the energy and three frequency bands are scored. A conservative estimate is made based on the score, and if it is too high, the number of broken wires is determined.

2. The method for identifying broken wires in parallel steel wire cables based on acoustic emission signals according to claim 1, characterized in that, In the feature fingerprint database, each number of broken wires corresponds to an energy ratio range and a low-frequency energy percentage range.

3. The method for identifying broken wires in parallel steel wire cables based on acoustic emission signals according to claim 1, characterized in that, The feature fingerprint database specifically includes: The number of broken wires indicates the number of broken wires in a parallel steel wire cable, and the value is 1, 2, 3, or 4. The energy ratio range is used to represent the range of values ​​for the energy ratio of the acoustic emission signal with broken wires. It is divided into [0.8, 1.2], [2.0, 2.7], [3.1, 4.5] and [4.4, 6.8] according to the different numbers of broken wires. The low-frequency energy percentage range is used to represent the range of values ​​for the low-frequency energy percentage of the broken wire acoustic emission signal. It is divided into [67.6%, 74.4%], [78.6%, 81.7%], [84.2%, 86.5%] and [89.1%, 91.2%] according to the different number of broken wires. The intermediate frequency energy ratio range is used to represent the range of values ​​for the intermediate frequency energy ratio of the broken wire acoustic emission signal. It is divided into [24.7%, 31.3%], [17.7%, 20.8%], [13.1%, 15.3%] and [8.3%, 10.4%] according to the different number of broken wires. The high-frequency energy percentage range is used to represent the range of values ​​for the high-frequency energy percentage of the broken wire acoustic emission signal. It is divided into [0.728%, 0.958%], [0.535%, 0.583%], [0.399%, 0.464%] and [0.355%, 0.364%] according to the different number of broken wires. The characteristic fingerprint ID is used to name the different numbers of broken wires in parallel steel wire cables. The names are F1, F2, F3, and F4.

4. The method for identifying broken wires in parallel steel wire cables based on acoustic emission signals according to claim 1, characterized in that, The low-frequency band ranges from 0 to 125 kHz, the mid-frequency band ranges from 156.25 to 250 kHz, and the high-frequency band ranges from 343.75 to 500 kHz.

5. The method for identifying broken wires in parallel steel wire cables based on acoustic emission signals according to claim 1, characterized in that, The method for calculating the energy proportion of each frequency band is as follows: the energy proportion of each frequency band is calculated by performing a 5-level wavelet packet decomposition using the Db5 wavelet basis function.

6. The method for identifying broken wires in parallel steel wire cables based on acoustic emission signals according to claim 1, characterized in that, The process of setting up the acoustic emission sensor is as follows: the acoustic emission sensor is placed at the center of the uppermost cross section of the parallel steel wire cable anchoring end.

7. The method for identifying broken wires in parallel steel wire cables based on acoustic emission signals according to claim 1, characterized in that, The scoring mechanism is as follows: a total score of 100 points, of which energy accounts for 40 points, and high frequency, mid frequency, and low frequency each account for 20 points. The scoring mechanism specifically includes: energy ratio range, low-frequency energy proportion range, mid-frequency energy proportion range, high-frequency energy proportion range, corresponding root tendency, and score. The corresponding number of wires is used to indicate the number of parallel wire cables with a strong tendency to break, with values ​​of 1, 2, 3, or 4 wires with a strong tendency. The score is used to evaluate different numbers of broken wires. The score is calculated based on different energy ratio ranges and the corresponding wire number tendency. When it is in the transition range, the score is given proportionally.

Images