A method and system for nondestructive detection and identification of wood borers

Abstract

The invention discloses a nondestructive detection and identification method and system for wood borers, and specifically relates to the technical field of wood detection. The method divides wood into a number of detection areas, dynamically adjusts the sound wave frequency and emission intensity according to the density and thickness of each area, designs a reasonable sound wave transmission path and detection point position, records the energy attenuation characteristics and spectrum amplitude fluctuation characteristics of the reflected signal, assigns a weight to the sound wave interference degree of the detection area, generates a comprehensive detection accuracy index, compares the accuracy index with a threshold, divides the detection result into two categories of accuracy and inaccuracy, and dynamically adjusts the detection parameters according to the detection abnormality degree in the inaccuracy area, effectively solves the problems of false alarm and missed alarm caused by the interference of the natural characteristics of wood in traditional sound wave detection, improves the detection accuracy and reliability, reduces the dependence on other supplementary means, and significantly reduces the detection cost.

Description

Technical Field

[0001] The invention relates to the technical field of wood detection, and in particular to a method and system for nondestructive detection and identification of wood borers. Background Art

[0002] Nondestructive testing and identification of wood borers is a technical method used to detect and identify wood borer infestations without destroying the wood structure. Traditional wood borer detection usually requires drilling or cutting into the wood, which may affect the strength and integrity of the wood. Nondestructive testing uses non-destructive means such as sound waves, infrared imaging, ultrasound, and X-rays to detect wood borers or damaged areas inside the wood. The identification system analyzes the test data to determine whether wood borers are present and their distribution, thereby helping to promptly detect and deal with wood borer problems and improve the service life and safety of the wood.

[0003] Traditional methods for detecting wood borers have many limitations, especially acoustic wave detection technology, which faces several technical problems and interference factors in application, affecting the accuracy and reliability of detection. Existing acoustic wave detection methods are usually significantly interfered by the natural characteristics of wood. For example, humidity, density changes, knots, etc. will interfere with the propagation of sound waves, making it difficult for the reflected signal energy attenuation characteristics and spectrum amplitude fluctuation characteristics to directly reflect the wood borer infestation. This interference leads to false alarms or missed alarms in detection, limiting the effectiveness of acoustic wave detection in practical applications. Especially in wood with high humidity and complex structure, the propagation speed and energy attenuation characteristics of sound waves in different density areas are different, making it difficult for existing detection methods to accurately identify the presence and distribution of wood borers. At the same time, due to the low accuracy of the test results, it is necessary to rely on other supplementary detection methods for secondary verification, which increases the cost and complexity of detection. In addition, the sound wave frequency and emission intensity in the existing technology are often set fixed, and the continuous impact of the dynamic changes of environmental factors on the test results is not considered, making it difficult to effectively respond to the detection needs in complex environments. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system for nondestructive detection and identification of wood borers to overcome the shortcomings of the background technology.

[0005] In order to achieve the above object, the present invention provides the following technical solution: a method for nondestructive detection and identification of wood borers, comprising the following steps:

[0006] S1: Divide the wood into several detection areas, and determine the sound wave frequency and emission intensity to be used according to the density and thickness of the wood in each detection area, and determine the sound wave transmission path and detection point location according to the shape and size of the wood;

[0007] S2: emitting sound waves longitudinally and transversely at the detection point position in the detection area, and recording basic data of the sound wave propagation in the wood, the basic data including the energy attenuation characteristics of the reflected signal of the sound wave and the signal spectrum amplitude fluctuation characteristics;

[0008] S3: Determine the weight assignment of the acoustic wave detection interference degree in each detection area according to the reflected signal energy attenuation characteristics and the signal spectrum amplitude fluctuation characteristics, and obtain the wood borer detection accuracy index by weighted average calculation of the weight assignment of the acoustic wave detection interference degree in each detection area;

[0009] S4: After comparing and analyzing the wood borer detection accuracy index with the detection accuracy threshold, the wood borer detection results are divided into accurate detection results and inaccurate detection results;

[0010] S5: For wood with inaccurate detection results, the sound wave frequency and emission intensity used in the detection area are dynamically adjusted according to the abnormal degree of wood borer detection accuracy within a fixed time period.

[0011] Preferably, in S2, after analyzing the energy attenuation characteristics of the sound wave reflected signal in the wood, a reflected signal energy attenuation abnormality index is generated, and the method for obtaining the reflected signal energy attenuation abnormality index is:

[0012] The sound waves are emitted in the detection area of ​​the wood and the reflected signals are recorded to obtain a time series data set x(s), where s represents time. The reflected signal x(s) is Fourier transformed to convert it into a frequency domain signal X(f), that is: ; Where X(f) is the complex signal in the frequency domain, f is the frequency, Represents the rotation factor in complex form, converting the frequency domain signal X(f) into an energy spectrum E(f) to represent the energy intensity of each frequency component. The calculation formula of the energy spectrum is: ; Perform the same frequency domain analysis on intact wood samples without borer infestation or damage to obtain a reference energy spectrum , for the actual energy spectrum E(f) in the detection area and the reference energy spectrum , calculate the energy attenuation coefficient A(f) at each frequency, the expression is: ; If E(f) is less than , calculate the reflected signal energy attenuation anomaly index within the entire spectrum range. The specific calculation formula is: ; where N is the number of frequency points within the effective frequency range, and are the lowest and highest frequencies in the analysis frequency range respectively, and SF is the abnormal index of reflected signal energy attenuation.

[0013] Preferably, in S2, after analyzing the signal spectrum amplitude fluctuation characteristics of the sound wave in the wood, a signal spectrum amplitude fluctuation index is generated, and the method for obtaining the signal spectrum amplitude fluctuation index is:

[0014] Obtain the reflected signal data x(t) of the wood detection area, and use EMD to decompose the signal x(t) into several intrinsic mode functions IMFs. Each IMF represents a different frequency component of the signal. The signal x(t) is decomposed into a series of and a residual , the expression is: ; where n is the number of IMFs, is the i-th IMF component; for each IMF component After Hilbert transform, we get ,Will and its Hilbert transform Combined into complex form analytical signals , the expression is: ; In the formula, is the instantaneous amplitude, is the instantaneous phase, instantaneous frequency The expression is obtained by differentiating the instantaneous phase with time: ; Analyze the instantaneous amplitude of each IMF , the fluctuation amplitude is calculated by the difference between the maximum and minimum amplitudes , the expression is: ; Perform weighted averaging on the fluctuation amplitudes of all IMF components to generate the overall signal spectrum amplitude fluctuation index.

[0015] Preferably, in S3, the weight assignment of the sound wave detection interference degree in each detection area is determined according to the reflected signal energy attenuation characteristics and the signal spectrum amplitude fluctuation characteristics, specifically:

[0016] The reflected signal energy attenuation anomaly index and the signal spectrum amplitude fluctuation index are converted into feature vectors, and the feature vectors are used as the input of the machine learning model. The machine learning model uses the weighted assignment labels of the acoustic detection interference degree in the detection area predicted by each group of feature vectors as the prediction target, and takes minimizing the sum of the prediction errors of the weighted assignment labels of the acoustic detection interference degree in all detection areas as the training target. The machine learning model is trained until the sum of the prediction errors converges and the model training is stopped. The weighted assignment of the acoustic detection interference degree in the detection area is determined according to the model output results, wherein the machine learning model is a polynomial regression model, and the weighted average calculation of the weighted assignment of the acoustic detection interference degree in each detection area is used to obtain the wood borer detection accuracy index.

[0017] Preferably, in S4, after comparing and analyzing the wood borer detection accuracy index with the detection accuracy threshold, the wood borer detection result is divided into an accuracy detection result and an inaccuracy detection result, specifically:

[0018] The acquired wood borer detection accuracy index is compared with the detection accuracy threshold. If the wood borer detection accuracy index is greater than or equal to the detection accuracy threshold, it means that the accuracy of wood borer detection is high. At this time, a high-accuracy detection signal is generated, and the detection result of wood borer is classified as an accurate detection result. If the wood borer detection accuracy index is less than the detection accuracy threshold, it means that the accuracy of wood borer detection is low. At this time, a low-accuracy detection signal is generated, and the detection result of wood borer is classified as an inaccurate detection result, and the wood is marked.

[0019] Preferably, in S5, for wood with inaccurate detection results, the sound wave frequency and emission intensity used in the detection area are dynamically adjusted according to the abnormal degree of wood borer detection accuracy within a fixed time period, specifically:

[0020] Assume that the abnormal degree of detection accuracy is , which is used to measure the deviation between the wood borer detection accuracy index and the accuracy threshold. The calculation expression is: ; In the formula, the accuracy threshold is the preset minimum detection accuracy standard. When the deviation is greater than the reference threshold, the sound wave frequency and emission intensity are adjusted;

[0021] By adjusting the sound wave frequency f and the emission intensity I, the accuracy of the detection result can be optimized. is the adjusted frequency, the original frequency is ,according to Dynamically adjust frequency: ; where α is the adjustment coefficient, set is the adjusted emission intensity, and the original intensity is ,according to Dynamically adjust emission intensity: ; Where β is the strength adjustment coefficient;

[0022] In each fixed time period, according to the calculated Update the values ​​of the sound wave frequency f and the emission intensity I to form a dynamic adjustment sequence. After adjusting the frequency and emission intensity, re-test, update the test data, compare the new test accuracy index with the accuracy threshold, and verify the adjustment effect; if the degree of deviation When the deviation is still greater than the reference threshold, the frequency and intensity continue to be adjusted.

[0023] The present invention also provides a nondestructive detection and identification system for wood borers, comprising a detection area division module, an acoustic signal acquisition module, an accuracy index calculation module, an accuracy classification module and a parameter dynamic adjustment module;

[0024] Detection area division module: divides the wood into several detection areas, and determines the sound wave frequency and emission intensity to be used according to the density and thickness of the wood in each detection area, and determines the sound wave transmission path and detection point location according to the shape and size of the wood;

[0025] Acoustic signal acquisition module: emits acoustic waves longitudinally and transversely at the detection point position within the detection area, and records the basic data of the propagation of acoustic waves in the wood, including the energy attenuation characteristics of the reflected signal of the acoustic wave and the amplitude fluctuation characteristics of the signal spectrum;

[0026] Accuracy index calculation module: Determine the weight assignment of the acoustic wave detection interference degree in each detection area according to the reflected signal energy attenuation characteristics and the signal spectrum amplitude fluctuation characteristics, and obtain the wood borer detection accuracy index by weighted average calculation of the weight assignment of the acoustic wave detection interference degree in each detection area;

[0027] Accuracy classification module: after comparing and analyzing the wood borer detection accuracy index with the detection accuracy threshold, the wood borer detection results are divided into accurate detection results and inaccurate detection results;

[0028] Dynamic parameter adjustment module: For wood with inaccurate detection results, the sound wave frequency and emission intensity used in the detection area are dynamically adjusted according to the abnormal degree of wood borer detection accuracy within a fixed time period.

[0029] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0030] 1. The present invention improves the accuracy and reliability of wood borer detection by dynamically adjusting the acoustic wave detection parameters, and effectively solves the problem of false alarms or missed alarms caused by interference from natural characteristics (such as humidity, density, and knots) in traditional acoustic wave detection technology. By dividing the wood into detection areas, the acoustic wave frequency, emission intensity, transmission path, and detection point settings in the detection area are optimized in combination with density and thickness characteristics, thereby minimizing the impact of environmental characteristics on the detection results. At the same time, by recording and analyzing the energy attenuation characteristics of the reflected signal and the amplitude fluctuation characteristics of the signal spectrum, a weighted value for the degree of detection interference is generated, and the accuracy index of wood borer detection is comprehensively calculated, and the results of the detection area are automatically classified.

[0031] 2. The present invention uses a machine learning model for dynamic adjustment, optimizing the sound wave frequency and emission intensity in real time according to the degree of abnormality in detection accuracy. Through this adaptive adjustment strategy, the detection system can achieve continuous high-precision detection in different environments and complex wood structures, reduce dependence on other detection methods, reduce detection costs, and improve overall detection efficiency, making it more suitable for real-time monitoring and maintenance applications of complex or large wood structures. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0033] Figure 1 The figure is a flow chart of the method of the present invention.

[0034] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0036] Example 1, please refer to Figure 1 and Figure 2 As shown, the nondestructive detection and identification method of wood borers described in this embodiment includes the following steps:

[0037] S1: Divide the wood into several detection areas, and determine the sound wave frequency and emission intensity to be used according to the density and thickness of the wood in each detection area, and determine the sound wave transmission path and detection point location according to the shape and size of the wood;

[0038] S2: emitting sound waves longitudinally and transversely at the detection point position in the detection area, and recording basic data of the sound wave propagation in the wood, the basic data including the energy attenuation characteristics of the reflected signal of the sound wave and the signal spectrum amplitude fluctuation characteristics;

[0039] S3: Determine the weight assignment of the acoustic wave detection interference degree in each detection area according to the reflected signal energy attenuation characteristics and the signal spectrum amplitude fluctuation characteristics, and obtain the wood borer detection accuracy index by weighted average calculation of the weight assignment of the acoustic wave detection interference degree in each detection area;

[0040] S4: After comparing and analyzing the wood borer detection accuracy index with the detection accuracy threshold, the wood borer detection results are divided into accurate detection results and inaccurate detection results;

[0041] S5: For wood with inaccurate detection results, the sound wave frequency and emission intensity used in the detection area are dynamically adjusted according to the abnormal degree of wood borer detection accuracy within a fixed time period.

[0042] In S1, the wood is divided into several detection areas, and the sound wave frequency and emission intensity used are determined according to the density and thickness of the wood in each detection area, and the sound wave transmission path and detection point position are determined according to the shape and size of the wood, specifically:

[0043] Evaluate whether multiple inspection areas are needed based on the overall shape, size and use of the wood. Larger wood structures are often divided into multiple small areas for more accurate inspection. Define the boundaries of each inspection area based on the size of the wood, structural complexity and coverage of the inspection equipment to ensure that all areas are of appropriate size, evenly covered and reflect the different characteristics of the wood (such as density and thickness variations). Assign numbers to each inspection area and mark them for subsequent data collection, processing and analysis. Clear area numbers help record and manage inspection data.

[0044] Measure or estimate the density of each area by non-destructive means (e.g., mass-to-volume-based calculations or using preliminary sound wave reflection data). Density variations are often caused by natural wood characteristics (e.g., knots or moisture changes) and need to be documented. Use a thickness gauge or ultrasonic thickness meter to measure the thickness of the wood at each test area, especially for wood with uneven thickness, and obtain an average thickness by measuring multiple points. Record the density and thickness data for each test area and mark it with a diagram for reference in subsequent steps to determine the appropriate sound wave parameters.

[0045] Select the appropriate sound wave frequency according to the density and thickness characteristics of each area. Generally, high-density areas require higher-frequency sound waves for detection to more clearly reflect the internal structure, while thicker areas may require lower-frequency sound waves to increase penetration depth. Frequency selection example: High-frequency sound waves (such as 20-40 kHz) can be selected for areas with low density and small thickness, while low-frequency sound waves (such as 5-20 kHz) are suitable for areas with high density and large thickness. The emission intensity is closely related to the thickness of the wood. A higher emission intensity is required for areas with larger thickness to ensure that the sound waves can penetrate the entire area and be detected by the receiver; while a moderate emission intensity is used for areas with smaller thickness to avoid signal reflection interference caused by excess energy. Emission intensity adjustment example: For example, a low-intensity emission can be set for areas below 20 cm, and the intensity can be increased (such as 2 times or more) for areas with a thickness greater than 20 cm.

[0046] Identify natural characteristics such as knots, density changes, and moisture differences within the inspection area. These factors affect the propagation path of the sound waves, so structural analysis is needed to avoid interference areas as much as possible or adjust the path. Determine the best transmission path for the sound waves based on the shape, density, and thickness of each area. For example, longitudinal transmission can better detect longer areas, while transverse transmission is suitable for evaluating areas with varying widths and thicknesses. If the thickness and density of the wood are unevenly distributed, multiple paths can be added to better cover the area. For example, design multiple longitudinal paths in areas with large thickness variations on long wooden boards, and add transverse paths in areas with varying widths.

[0047] Place transmitters and receivers at the start and end of each sound wave path, and avoid placing detection points in knots or high humidity areas to reduce the interference of natural characteristics on the sound wave signal. If there are points with significant changes in density or thickness within the area, auxiliary detection points can be placed at these locations to receive and analyze mid-way signals to improve detection accuracy. Mark the specific location of each detection point clearly on the wood surface to ensure accuracy during subsequent testing, and record it in the area detection chart.

[0048] Record the sound wave frequency, emission intensity, path layout and detection point location of all areas in a table to ensure consistency in each test. After the record is completed, review whether the parameter settings of each area are reasonable, especially whether the interference of natural characteristics on the test results can be effectively avoided. If necessary, further adjustments can be made to the frequency, emission intensity, path or detection point location.

[0049] S2: The sound waves are emitted longitudinally and transversely at the detection point positions within the detection area, and the basic data of the sound waves propagating in the wood are recorded, wherein the basic data include the energy attenuation characteristics of the reflected signal of the sound waves and the amplitude fluctuation characteristics of the signal spectrum.

[0050] At the predetermined detection point of each inspection area, place the acoustic wave transmitter and receiver in the longitudinal and transverse directions respectively. The arrangement of the transmitter and receiver should ensure that the inner area of ​​the wood is covered so that sufficient information can be received. Check the frequency and emission intensity settings of the transmitter to ensure that they meet the acoustic wave parameters previously determined for the inspection area to avoid signals that are too strong or too weak and affect the data quality.

[0051] The sound waves are emitted in the longitudinal direction to ensure that the sound waves penetrate the entire detection area. The propagation path of longitudinal sound waves is usually longer and is suitable for detecting the length of the wood. The sound waves are emitted in the transverse direction. The transverse detection path usually covers the width or thickness area of ​​the wood to capture a wider range of structural information. If the wood structure is complex, the emission angle can be adjusted to avoid obvious knots, cracks or areas with high humidity, thereby reducing the interference of natural characteristics on the signal.

[0052] The receiver captures the sound waves reflected in the wood. Uneven structures or areas of woodworm damage can cause the sound waves to attenuate significantly during propagation. The initial energy of the incident signal is compared with the energy of the reflected signal, and the energy attenuation ratio is calculated. The attenuation characteristic is usually expressed as: ; For areas with borer infestation or uneven density, the attenuation rate of the reflected signal is usually higher. The attenuation characteristics are recorded as numerical data for subsequent analysis of the structural integrity of the area.

[0053] After analyzing the energy attenuation characteristics of the sound wave reflected signal in the wood, the reflected signal energy attenuation abnormality index is generated. The method for obtaining the reflected signal energy attenuation abnormality index is as follows:

[0054] The sound waves are emitted in the detection area of ​​the wood and the reflected signals are recorded to obtain a time series data set x(s), where s represents time. It is ensured that the collected signal has a sufficiently high sampling frequency fs to satisfy the sampling theorem (i.e. ,in x is the highest frequency contained in the signal).

[0055] Perform Fourier transform on the reflected signal x(s) and convert it into a frequency domain signal X(f), that is: ; Where X(f) is the complex signal in the frequency domain, f is the frequency, Represents the rotation factor in complex form, converting the frequency domain signal X(f) into an energy spectrum E(f) to represent the energy intensity of each frequency component. The calculation formula of the energy spectrum is: ; Perform the same frequency domain analysis on intact wood samples without borer infestation or damage to obtain a reference energy spectrum , for the actual energy spectrum E(f) in the detection area and the reference energy spectrum , calculate the energy attenuation coefficient A(f) at each frequency, the expression is: ; If E(f) is less than , then the A(f) value is closer to 1, indicating stronger energy attenuation. Calculate the energy attenuation anomaly index of the reflected signal in the entire spectrum range. The specific calculation formula is: ; where N is the number of frequency points within the effective frequency range, and are the lowest and highest frequencies in the analysis frequency range respectively, and SF is the abnormal index of reflected signal energy attenuation.

[0056] When the reflected signal energy attenuation anomaly index is larger, it indicates that the reflected signal energy in the detection area is more significantly attenuated than the reference energy spectrum of intact wood, which means that the sound wave is more disturbed when propagating in this area. Generally, a larger anomaly index indicates that there are more uneven structures or gaps inside the wood, such as woodworm damage, holes or cracks. At this time, the energy of the sound wave is continuously dissipated during propagation, especially in high-density defects or damaged areas, where the reflection increases significantly and the attenuation degree increases. A high anomaly index is usually a warning sign of woodworm infestation or structural damage, indicating that the detection accuracy of the area may be affected and requires further analysis and confirmation.

[0057] On the contrary, when the reflected signal energy attenuation anomaly index is smaller, it indicates that the reflected signal energy in the detection area is closer to the reference energy spectrum of intact wood, the attenuation during sound wave propagation is relatively small, and there is less interference. This usually means that the wood structure is uniform and dense, there are no obvious defects or damage inside, and the possibility of woodworm infestation is low. A small anomaly index usually indicates that the signal attenuation and reflection in the detection area are normal, and the internal environment of the wood has little interference with sound wave propagation. At this time, the reliability and accuracy of the test results are high, and the structural integrity of the area is good.

[0058] After analyzing the amplitude fluctuation characteristics of the signal spectrum of the sound wave in the wood, the signal spectrum amplitude fluctuation index is generated. The method for obtaining the signal spectrum amplitude fluctuation index is:

[0059] Get the reflected signal data x(t) of the wood detection area, apply EMD to decompose the signal x(t) into several intrinsic mode functions IMF, each IMF represents a different frequency component of the signal, find all local maxima and minima from the signal, use the extreme points to build the upper envelope and lower envelope, calculate the mean of the upper and lower envelopes, subtract the mean from the signal to get a new signal, and repeat the above process until the IMF condition is met. Separate the obtained IMF components from the signal, repeat the above steps until all IMFs are decomposed, and the signal x(t) is decomposed into a series of and a residual , the expression is: ; where n is the number of IMFs, is the i-th IMF component; for each IMF component After Hilbert transform, we get ,Will and its Hilbert transform Combined into complex form analytical signals , the expression is: ; In the formula, is the instantaneous amplitude, is the instantaneous phase, instantaneous frequency The expression is obtained by differentiating the instantaneous phase with time: ; Analyze the instantaneous amplitude of each IMF , the fluctuation amplitude is calculated by the difference between the maximum and minimum amplitudes , the expression is: ; Perform weighted averaging on the fluctuation amplitudes of all IMF components to generate the overall signal spectrum amplitude fluctuation index.

[0060] When the signal spectrum amplitude fluctuation index is larger, it indicates that the spectrum amplitude in the detection area fluctuates significantly, and the amplitude of the signal varies greatly at different frequencies and time periods. This usually means that the internal structure of the wood is uneven, and there are defects such as cavities, cracks, or woodworms. These defects lead to significant changes in sound wave reflection and attenuation, which intensifies the fluctuation of the spectrum amplitude. Therefore, the larger the fluctuation index, the higher the degree of interference with the sound wave detection, the more serious damage or woodworms may exist in the detection area, the lower the consistency and stability of the signal propagation process, and the higher the difficulty of detection.

[0061] On the contrary, the smaller the signal spectrum amplitude fluctuation index, the more stable the signal spectrum amplitude is, indicating that the wood structure in the detection area is uniform and dense. A small fluctuation index usually means that there are no significant cavities or defects in the detection area, the possibility of woodworm infestation or structural damage is low, and the sound wave propagation in the area is relatively stable and less disturbed. Such a small fluctuation index reflects that the signal spectrum amplitude is highly consistent, and the sound wave detection in the detection area is less disturbed, so the accuracy and reliability of the detection results are higher.

[0062] S3: Determine the weighted value of the acoustic wave detection interference degree in each detection area according to the reflected signal energy attenuation characteristics and the signal spectrum amplitude fluctuation characteristics, and obtain the wood borer detection accuracy index by weighted average calculation of the weighted value of the acoustic wave detection interference degree in each detection area.

[0063] The reflected signal energy attenuation anomaly index and the signal spectrum amplitude fluctuation index are converted into feature vectors, and the feature vectors are used as the input of the machine learning model. The machine learning model uses the weighted assignment labels of the acoustic detection interference degree in the detection area predicted by each group of feature vectors as the prediction target, and takes minimizing the sum of the prediction errors of the weighted assignment labels of the acoustic detection interference degree in all detection areas as the training target. The machine learning model is trained until the sum of the prediction errors converges and the model training is stopped. The weighted assignment of the acoustic detection interference degree in the detection area is determined according to the model output results, wherein the machine learning model is a polynomial regression model, and the weighted average calculation of the weighted assignment of the acoustic detection interference degree in each detection area is used to obtain the wood borer detection accuracy index.

[0064] The method for obtaining the weighted value of the acoustic wave detection interference degree in the detection area is as follows: from the first eigenvector training data of the trained machine learning model, the corresponding function expression is obtained: ; In the formula, F is the output function of the model, SF is the abnormal index of reflected signal energy attenuation, GH is the signal spectrum amplitude fluctuation index, and CG is the weight assignment of the degree of interference of acoustic wave detection in the detection area.

[0065] S4: After comparing and analyzing the wood borer detection accuracy index with the detection accuracy threshold, the wood borer detection results are divided into accurate detection results and inaccurate detection results.

[0066] The acquired wood borer detection accuracy index is compared with the detection accuracy threshold. If the wood borer detection accuracy index is greater than or equal to the detection accuracy threshold, it means that the accuracy of wood borer detection is high. At this time, a high-accuracy detection signal is generated, and the detection result of wood borer is classified as an accurate detection result. If the wood borer detection accuracy index is less than the detection accuracy threshold, it means that the accuracy of wood borer detection is low. At this time, a low-accuracy detection signal is generated, and the detection result of wood borer is classified as an inaccurate detection result, and the wood is marked.

[0067] S5: For wood with inaccurate detection results, the sound wave frequency and emission intensity used in the detection area are dynamically adjusted according to the abnormal degree of wood borer detection accuracy within a fixed time period.

[0068] Assume that the abnormal degree of detection accuracy is , which is used to measure the deviation between the wood borer detection accuracy index and the accuracy threshold. The calculation expression is: ; In the formula, the accuracy threshold is the preset minimum detection accuracy standard. When it is greater than the deviation reference threshold, it indicates that the sound wave frequency or emission intensity in the detection area needs to be adjusted.

[0069] By adjusting the sound wave frequency f and the emission intensity I, the accuracy of the detection result can be optimized. is the adjusted frequency, the original frequency is ,according to Dynamically adjust frequency: ; where α is the adjustment coefficient, which is used to control the sensitivity of frequency changes. A higher α value will result in a larger frequency adjustment, and the typical range is 0.1≤α≤0.5. When it is larger, the frequency increase or decrease will be larger, which helps to improve the penetration or sensitivity of the signal in high interference conditions.

[0070] set up is the adjusted emission intensity, and the original intensity is ,according to Dynamically adjust emission intensity: ; Where β is the intensity adjustment coefficient, which determines the response of the emission intensity to the abnormality, and is usually set to 0.1≤β≤0.3. When the transmission intensity is large, the transmission intensity will be significantly improved, thereby enhancing the signal penetration ability and improving the quality of the reflected signal.

[0071] In each fixed time period, according to the calculated Update the values ​​of the sound wave frequency f and the emission intensity I to form a dynamic adjustment sequence. After adjusting the frequency and emission intensity, re-test and update the test data. Compare the new test accuracy index with the accuracy threshold to verify the adjustment effect. If the degree of deviation If the value is still greater than the deviation reference threshold, continue to adjust the frequency and intensity until the detection accuracy reaches the ideal level.

[0072] In this embodiment, the wood is first divided into multiple detection areas, and the appropriate sound wave frequency, emission intensity, sound wave transmission path and detection point position are determined according to the density and thickness in each area. Then, sound waves are emitted at the detection point position from the longitudinal and transverse directions, and the energy attenuation characteristics of the reflected signal and the spectrum amplitude fluctuation characteristics of the sound wave propagation are recorded. Based on these characteristics, the degree of interference of the sound wave detection in each detection area is evaluated, the corresponding weight assignment is assigned, and the weight assignment of all detection areas is weighted averaged to obtain the wood borer detection accuracy index. This accuracy index is compared with the set detection accuracy threshold, and the detection results are divided into two categories: high accuracy and low accuracy. For inaccurate detection areas, the sound wave frequency and emission intensity are dynamically adjusted according to the abnormal degree of detection accuracy to optimize the detection parameters and improve the detection reliability.

[0073] Embodiment 2, a wood borer nondestructive detection and identification system described in this embodiment includes a detection area division module, an acoustic signal acquisition module, an accuracy index calculation module, an accuracy classification module and a parameter dynamic adjustment module;

[0074] Detection area division module: divides the wood into several detection areas, and determines the sound wave frequency and emission intensity to be used according to the density and thickness of the wood in each detection area, and determines the sound wave transmission path and detection point location according to the shape and size of the wood;

[0075] Acoustic signal acquisition module: emits acoustic waves longitudinally and transversely at the detection point position within the detection area, and records the basic data of the propagation of acoustic waves in the wood, including the energy attenuation characteristics of the reflected signal of the acoustic wave and the amplitude fluctuation characteristics of the signal spectrum;

[0076] Accuracy index calculation module: Determine the weight assignment of the acoustic wave detection interference degree in each detection area according to the reflected signal energy attenuation characteristics and the signal spectrum amplitude fluctuation characteristics, and obtain the wood borer detection accuracy index by weighted average calculation of the weight assignment of the acoustic wave detection interference degree in each detection area;

[0077] Accuracy classification module: after comparing and analyzing the wood borer detection accuracy index with the detection accuracy threshold, the wood borer detection results are divided into accurate detection results and inaccurate detection results;

[0078] Dynamic parameter adjustment module: For wood with inaccurate detection results, the sound wave frequency and emission intensity used in the detection area are dynamically adjusted according to the abnormal degree of wood borer detection accuracy within a fixed time period.

[0079] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0080] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0081] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0082] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A method for nondestructive detection and identification of wood borers, characterized in that: The steps include: S1: Divide the wood into several detection areas, and determine the sound wave frequency and emission intensity to be used according to the density and thickness of the wood in each detection area, and determine the sound wave transmission path and detection point location according to the shape and size of the wood; S2: emitting sound waves longitudinally and transversely at the detection point position in the detection area, and recording basic data of the sound wave propagation in the wood, the basic data including the energy attenuation characteristics of the reflected signal of the sound wave and the signal spectrum amplitude fluctuation characteristics; S3: Determine the weighted value of the acoustic wave detection interference degree in each detection area according to the reflected signal energy attenuation characteristics and the signal spectrum amplitude fluctuation characteristics, and obtain the wood borer detection accuracy index by weighted average calculation of the weighted values ​​of the acoustic wave detection interference degree in each detection area, specifically: convert the reflected signal energy attenuation anomaly index and the signal spectrum amplitude fluctuation index into feature vectors, and use the feature vectors as inputs of the machine learning model. The machine learning model uses each group of feature vectors to predict the weighted value labels of the acoustic wave detection interference degree in the detection area as the prediction target, and uses minimizing the sum of the prediction errors of the weighted value labels of the acoustic wave detection interference degree in all detection areas as the training target. The machine learning model is trained until the sum of the prediction errors converges and the model training is stopped. The weighted value of the acoustic wave detection interference degree in the detection area is determined according to the model output results, wherein the machine learning model is a polynomial regression model, and the weighted average calculation of the weighted value of the acoustic wave detection interference degree in each detection area is used to obtain the wood borer detection accuracy index; S4: After comparing and analyzing the wood borer detection accuracy index with the detection accuracy threshold, the wood borer detection results are divided into accurate detection results and inaccurate detection results; S5: For wood with inaccurate detection results, the sound wave frequency and emission intensity used in the detection area are dynamically adjusted according to the abnormal degree of wood borer detection accuracy within a fixed time period.

2. A method for nondestructive detection and identification of wood borers according to claim 1, characterized in that: In S2, the energy attenuation characteristics of the sound wave reflected signal in the wood are analyzed to generate the reflected signal energy attenuation abnormality index. The method for obtaining the reflected signal energy attenuation abnormality index is: The sound waves are emitted in the detection area of ​​the wood and the reflected signals are recorded to obtain a time series data set x(s), where s represents time. The reflected signal x(s) is Fourier transformed to convert it into a frequency domain signal X(f), that is: ; Where X(f) is the complex signal in the frequency domain, f is the frequency, Represents the rotation factor in complex form, converting the frequency domain signal X(f) into an energy spectrum E(f) to represent the energy intensity of each frequency component. The calculation formula of the energy spectrum is: ; Perform the same frequency domain analysis on intact wood samples without borer infestation or damage to obtain a reference energy spectrum , for the actual energy spectrum E(f) in the detection area and the reference energy spectrum , calculate the energy attenuation coefficient A(f) at each frequency, the expression is: ; If E(f) is less than , calculate the reflected signal energy attenuation anomaly index within the entire spectrum range. The specific calculation formula is: ; where N is the number of frequency points within the effective frequency range, and are the lowest and highest frequencies in the analysis frequency range respectively, and SF is the abnormal index of reflected signal energy attenuation.

3. A method for nondestructive detection and identification of wood borers according to claim 2, characterized in that: In S2, the signal spectrum amplitude fluctuation characteristics of the sound wave in the wood are analyzed to generate a signal spectrum amplitude fluctuation index. The method for obtaining the signal spectrum amplitude fluctuation index is: Obtain the reflected signal data x(t) of the wood detection area, and use EMD to decompose the signal x(t) into several intrinsic mode functions IMFs. Each IMF represents a different frequency component of the signal. The signal x(t) is decomposed into a series of and a residual , the expression is: ; where n is the number of IMFs, is the i-th IMF component; for each IMF component After Hilbert transform, we get ,Will and its Hilbert transform Combined into complex form analytical signals , the expression is: ; In the formula, is the instantaneous amplitude, is the instantaneous phase, instantaneous frequency The expression is obtained by differentiating the instantaneous phase with time: ; Analyze the instantaneous amplitude of each IMF , the fluctuation amplitude is calculated by the difference between the maximum and minimum amplitudes , the expression is: ; The fluctuation amplitudes of all IMF components are weighted averaged to generate the overall signal spectrum amplitude fluctuation index.

4. A method for nondestructive detection and identification of wood borers according to claim 1, characterized in that: In S4, after comparing and analyzing the wood borer detection accuracy index and the detection accuracy threshold, the wood borer detection results are divided into accurate detection results and inaccurate detection results, specifically: The acquired wood borer detection accuracy index is compared with the detection accuracy threshold. If the wood borer detection accuracy index is greater than or equal to the detection accuracy threshold, it means that the accuracy of wood borer detection is high. At this time, a high-accuracy detection signal is generated, and the detection result of wood borer is classified as an accurate detection result. If the wood borer detection accuracy index is less than the detection accuracy threshold, it means that the accuracy of wood borer detection is low. At this time, a low-accuracy detection signal is generated, and the detection result of wood borer is classified as an inaccurate detection result, and the wood is marked.

5. The method for nondestructive detection and identification of wood borers according to claim 1, characterized in that: In S5, for wood with inaccurate detection results, the sound wave frequency and emission intensity used in the detection area are dynamically adjusted according to the abnormal degree of wood borer detection accuracy within a fixed time period, specifically: Assume that the abnormal degree of detection accuracy is , which is used to measure the deviation between the wood borer detection accuracy index and the accuracy threshold. The calculation expression is: ; In the formula, the accuracy threshold is the preset minimum detection accuracy standard. When the deviation is greater than the reference threshold, the sound wave frequency and emission intensity are adjusted; By adjusting the sound wave frequency f and the emission intensity I, the accuracy of the detection result can be optimized. is the adjusted frequency, the original frequency is ,according to Dynamically adjust frequency: ; where α is the adjustment coefficient, set is the adjusted emission intensity, and the original intensity is ,according to Dynamically adjust emission intensity: ; Where β is the strength adjustment coefficient; In each fixed time period, according to the calculated Update the values ​​of the sound wave frequency f and the emission intensity I to form a dynamic adjustment sequence. After adjusting the frequency and emission intensity, re-test, update the test data, compare the new test accuracy index with the accuracy threshold, and verify the adjustment effect; if the degree of deviation When the deviation is still greater than the reference threshold, the frequency and intensity continue to be adjusted.

6. A non-destructive detection and identification system for wood borers, used to implement a non-destructive detection and identification method for wood borers according to any one of claims 1 to 5, characterized in that: It includes a detection area division module, an acoustic signal acquisition module, an accuracy index calculation module, an accuracy classification module and a parameter dynamic adjustment module; Detection area division module: divides the wood into several detection areas, and determines the sound wave frequency and emission intensity to be used according to the density and thickness of the wood in each detection area, and determines the sound wave transmission path and detection point location according to the shape and size of the wood; Acoustic signal acquisition module: emits acoustic waves longitudinally and transversely at the detection point position within the detection area, and records the basic data of the propagation of acoustic waves in the wood, including the energy attenuation characteristics of the reflected signal of the acoustic wave and the amplitude fluctuation characteristics of the signal spectrum; Accuracy index calculation module: Determine the weight assignment of the acoustic wave detection interference degree in each detection area according to the reflected signal energy attenuation characteristics and the signal spectrum amplitude fluctuation characteristics, and obtain the wood borer detection accuracy index by weighted average calculation of the weight assignment of the acoustic wave detection interference degree in each detection area; Accuracy classification module: after comparing and analyzing the wood borer detection accuracy index with the detection accuracy threshold, the wood borer detection results are divided into accurate detection results and inaccurate detection results; Dynamic parameter adjustment module: For wood with inaccurate detection results, the sound wave frequency and emission intensity used in the detection area are dynamically adjusted according to the abnormal degree of wood borer detection accuracy within a fixed time period.

Images