A method, system, device and medium for predicting mechanical properties of wood based on live tree stress wave data

Abstract

The present application relates to a kind of wood mechanical property prediction method, system, equipment and medium based on live standing tree stress wave data, wherein, method includes: obtaining live standing tree longitudinal wave velocity data, green density, live standing tree moisture content, forest tree germplasm and tree age;Live standing tree longitudinal wave velocity data is screened based on wave velocity gradient mutation response trigger knot defect, and moisture content is combined to implement correction, output standard longitudinal wave velocity data, and boundary effect compensation is executed, generates log equivalent wave velocity, gradient correction is carried out in combination with the radial spatial distribution relationship of detection node and pith, and output corrected wave velocity parameter;Through the coupling of multiple sources, dynamic elastic modulus prediction model is constructed, and dynamic elastic modulus is predicted and output, and is converted into static reference modulus, conditionally introduces tree age attenuation factor to exert correction, and obtains final static elastic modulus.The present application significantly improves the precision and applicability of live standing tree mechanical property prediction by coupling analysis and dynamic correction mechanism of multiple source stress wave data.

Description

Technical Field

[0001] This invention relates to the field of standing timber testing technology, and in particular to a method, system, device, and medium for predicting the mechanical properties of timber based on stress wave data of standing timber. Background Technology

[0002] As a natural composite material, the mechanical properties of wood determine its applicability and value in construction, furniture, and engineering. The static modulus of elasticity (MOE), often simply called the modulus of elasticity, is one of the most important mechanical properties, directly reflecting the wood's ability to resist elastic deformation. A higher MOE value indicates that the wood is less prone to bending and deformation, and its structural stability is stronger. Therefore, it plays a crucial role in timber structure design, timber grading, and performance evaluation of wood products. It is also an important early indicator for judging the quality of timber in forest genetic improvement and efficient breeding.

[0003] Existing techniques for measuring static modulus of elasticity mainly rely on felling standing trees and cutting standard specimens for laboratory mechanical testing. Because this method physically destroys the integrity of the wood, it inevitably leads to the depletion of forest resources and loss of ecological benefits, a problem that is particularly pronounced in applications involving precious tree species or ecologically sensitive areas.

[0004] In the field of standing tree testing, existing technologies typically rely on local wave velocities near the bark and sapwood to construct dynamic elastic modulus prediction models. However, wave velocities vary significantly across different radial locations of the wood. This leads to discrepancies between predictions based on local sapwood wave velocities and the actual overall mechanical properties of the wood. Such predictions are often only suitable for trend comparisons and cannot achieve quantitative assessments close to the measured values. Furthermore, differences in the biological characteristics of standing trees (including differences in tree germplasm and age stage) and the physical properties of the wood (such as dynamic changes in moisture content and internal knot defects) significantly interfere with the signal characteristics of the model's core parameter—stress wave velocity. Therefore, the prediction accuracy of existing methods is insufficient to meet the needs of practical engineering applications.

[0005] Furthermore, current elastic modulus prediction models are mostly developed for specific tree species or fixed age ranges. Due to the lack of a unified quantitative adaptation mechanism for the differences in wood structure among various tree species and the growth characteristics across different ages, existing technologies are unable to achieve elastic modulus prediction coverage for a wide range of tree species and standing trees of all ages, which seriously restricts the engineering and promotion value of the technology. Summary of the Invention

[0006] (a) Technical problems to be solved

[0007] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention provides a method for predicting the mechanical properties of wood based on stress wave data of living trees. It solves the technical problems of existing static elastic modulus measurement technology, which relies on destructive logging detection and causes resource and ecological conflicts; non-destructive testing of living trees, which only uses the local wave velocity of sapwood and causes the prediction results to deviate from the overall true value; insufficient accuracy of the prediction model due to the coupling interference of biological characteristics and physical properties of wood; and the limited universality of existing models due to the lack of quality adaptation mechanism across tree age and multiple tree species.

[0008] (II) Technical Solution

[0009] To achieve the above objectives, the main technical solutions adopted by the present invention include:

[0010] In a first aspect, embodiments of the present invention provide a method for predicting the mechanical properties of wood based on stress wave data of living trees, including:

[0011] By conducting physical measurements, biomarker analysis, and record association analysis on living trees, we obtained data on longitudinal wave velocity, green wood density, moisture content, tree germplasm, and tree age of living trees.

[0012] Based on the abrupt response of the wave velocity gradient, the longitudinal wave velocity data of the living trees is used to trigger the screening of knot defects, and dynamic correction is implemented in combination with the moisture content of the living trees to output standard longitudinal wave velocity data.

[0013] Boundary effect compensation is performed on the standard longitudinal wave velocity data according to the boundary effect parameters determined by the tree germplasm to generate the log equivalent wave velocity. Gradient correction is then performed in combination with the radial spatial distribution relationship between the detection nodes and the pith to output the corrected wave velocity parameters.

[0014] The corrected wave velocity parameter is coupled with the raw material density and the introduced density-wave velocity coupling coefficient through multiple sources to construct a dynamic elastic modulus prediction model and predict the output dynamic elastic modulus.

[0015] The dynamic elastic modulus is converted into a static baseline modulus based on the growth characteristics corresponding to the tree germplasm, and a continuously adjustable age decay factor is introduced according to the age condition to apply continuous decay correction, thus obtaining the final static elastic modulus.

[0016] Optionally, through physical measurements, biomarker analysis, and record association analysis of living trees, data on longitudinal wave velocity, greenwood density, moisture content, tree germplasm, and tree age of living trees can be obtained, including:

[0017] The mechanical guiding device is controlled to be deployed at a preset detection node position, penetrate the bark to reach a preset depth in the sapwood layer, and obtain the installation position of the stress wave sensor;

[0018] Based on the installation location of the stress wave sensor, guide the stress wave sensor to be deployed in the section of the standing tree with a diameter at breast height of 0.5-2.0m, according to the topology of equal spacing along the north-south axis of the trunk.

[0019] The standard impact hammer is controlled to perform vertical impact excitation, and the time series data stream of bidirectional stress wave propagation is acquired simultaneously from the north and south directions.

[0020] The sensor group is driven to repeatedly collect data streams N times. The effective propagation time is obtained by removing abnormal data with fluctuation amplitude exceeding 10%. The longitudinal wave velocity data of the living tree is obtained by combining the topological spacing of the sensors.

[0021] A growth cone was used to extract wood core samples from the diameter-to-breast-length section, and the mass and three-dimensional volume data of the wood core samples were collected simultaneously to determine the green wood density.

[0022] The moisture content is measured in real time by a pre-deployed moisture content sensor. After verifying the validity of the data, the current moisture content value is output. If a communication interruption or data drift is detected in the moisture content sensor, the corresponding moisture content in the preset green material moisture content parameter library is called as the current moisture content.

[0023] Read the identification code pre-set on the living tree to parse the forest tree germplasm. If the identification code is damaged, start the real-time comparison of near-infrared spectral features and DNA marker library to determine the forest tree germplasm.

[0024] The system calls up the forest stand archive to match the spatial coordinates of the current living trees. If historical records exist, the tree age is extracted directly; otherwise, the tree age is determined by generating inverse growth ring counts through core ring analysis.

[0025] Optionally, the longitudinal wave velocity data of living trees is used to trigger nodule defect screening based on the abrupt change in wave velocity gradient response, and dynamic correction is performed in conjunction with the moisture content of the living trees. The output standard longitudinal wave velocity data includes:

[0026] Wavelet packet decomposition was performed on the longitudinal wave velocity data of living trees to obtain the multi-band wave velocity feature components corresponding to each detection node;

[0027] The target component of the preset main frequency band in each node is selected from the multi-band wave velocity feature components, and wavelet threshold denoising is performed.

[0028] A synchronous squeezing wavelet transform is performed on the denoised target component signal to obtain the time spectrum aligned with adjacent detection nodes, the jump points on the time spectrum are detected, and the jump points on the time spectrum are corrected by interpolation.

[0029] Based on the corrected time spectrum, the wave velocity gradient sequence is obtained according to the longitudinal equidistant topological structure of the tree trunk, and mapped to a three-dimensional cylindrical coordinate system to generate a continuous wave velocity gradient matrix. The continuous wave velocity gradient matrix is ​​constructed based on the three-dimensional cylindrical coordinate system through longitudinal, circumferential and radial discretization, and integrates the propagation characteristics of stress waves in the wood fiber structure and the radial attenuation law as functions, which are used to characterize the gradient field where the wave velocity changes continuously with spatial position.

[0030] The longitudinal first derivative of the continuous wave velocity gradient matrix is ​​analyzed in real time. When the absolute value of the derivative between adjacent nodes exceeds the adaptive threshold determined by the current timber density and tree germplasm, it is automatically marked as a knot defect region, and the longitudinal wave velocity data of the living trees corresponding to the knot defect region is removed.

[0031] Based on the deviation between the current moisture content and the reference moisture content, and combined with the introduced moisture content correction coefficient, the longitudinal wave velocity data of the standing trees after removing the knot defect data is subjected to moisture content coupling compensation processing to obtain the primary corrected wave velocity.

[0032] For the primary correction wave velocity, standard longitudinal wave velocity data is generated by compensating for the anisotropic wave velocity distortion caused by the moisture absorption and expansion of wood fibers through nonlinear mapping.

[0033] Optionally, for the primary correction wave velocity, anisotropic wave velocity distortion caused by the hygroscopic expansion of wood fibers is compensated for through nonlinear mapping to generate standard longitudinal wave velocity data, including:

[0034] Based on the longitudinal gradient distribution of the primary correction wave velocity, the principal orientation angle of the living tree fiber is extracted by continuous wavelet transform;

[0035] Based on the offset between the current moisture content and the reference moisture content and the fiber principal orientation angle, a nonlinear compensation function matching the rheological properties of wood is dynamically constructed.

[0036] Based on the north-south orientation of the stress wave detection nodes, a differentiated orientation weighting factor is applied to the nonlinear compensation function; among which, the longitudinal propagation path weight is greater than the radial propagation path weight.

[0037] The primary correction wave velocity is input into the nonlinear compensation function to perform anisotropic distortion correction, generating standard longitudinal wave velocity data.

[0038] Optionally, boundary effect compensation is performed on the standard longitudinal wave velocity data according to the boundary effect parameters determined by the tree germplasm to generate the log equivalent wave velocity. Gradient correction is then performed based on the radial spatial distribution relationship between the detection nodes and the pith, and the output corrected wave velocity parameters include:

[0039] The standard longitudinal wave velocity data is converted from standing trees to logs using boundary effect parameters determined by tree germplasm to generate log equivalent wave velocities.

[0040] Based on the acquired north-south bidirectional stress wave propagation time series data stream, the stress wave propagation time difference of multiple detection nodes is extracted, and the propagation path of the stress wave around the spinal cord is dynamically reconstructed based on the propagation time difference.

[0041] The propagation path is discretized into multiple micro-segments with consistent fiber orientation. The main propagation direction of each micro-segment is determined, and a corresponding azimuth weighting factor is applied. The propagation time of each micro-segment is accumulated to obtain the theoretical propagation time of the stress wave to each node. By comparing it with the stress wave propagation time series data stream, a time difference residual matrix characterizing the spatial error distribution is generated.

[0042] Based on the spatial error distribution characteristics of the time difference residual matrix, the core coordinates are iteratively optimized until the preset convergence condition is met, and the optimal core 3D coordinates are output.

[0043] The radial spatial distance between the stress wave detection node and the core is analyzed based on the optimal three-dimensional coordinates of the core. The radial gradient correction factor is determined based on the radial spatial distribution relationship. The radial gradient correction factor increases as the node position approaches the sapwood region and decreases as it approaches the heartwood region.

[0044] The equivalent wave velocity of the log is calculated based on the radial gradient correction factor, and the output corrected wave velocity parameter is obtained by fusing boundary effect correction and radial gradient correction.

[0045] Optionally, the corrected wave velocity parameter is coupled with the raw material density and the introduced density-wave velocity coupling coefficient through multiple sources to construct a dynamic elastic modulus prediction model and predict the output dynamic elastic modulus, including:

[0046] Through regression analysis of multiple tree germplasm and tree age samples, the density-wave velocity coupling coefficient and elastic modulus compensation constant of each group were calculated in batches and stored in the dynamic parameter database.

[0047] Based on the current tree species and tree age, the corresponding density and wave velocity coupling coefficients and elastic modulus compensation constants are matched in real time from the dynamic parameter database.

[0048] The primary mechanical response value is generated by multiplying the square of the raw material density and the corrected wave velocity parameter. The primary mechanical response value is then multiplied by the current density and wave velocity coupling coefficient, and then superimposed with the current elastic modulus compensation constant to output the predicted value of the dynamic elastic modulus.

[0049] Optionally, the dynamic elastic modulus is converted into a static baseline modulus based on the growth characteristics corresponding to the tree germplasm, and a continuously adjustable age decay factor is introduced according to the age condition to apply continuous decay correction, resulting in the final static elastic modulus including:

[0050] Through regression analysis of multiple tree germplasm and tree age samples, the dynamic-static transition slope parameters and intercept parameters of each group were calculated in batches and stored in the static parameter database.

[0051] Based on the current tree species and tree age, extract the corresponding dynamic-static conversion slope parameters and intercept parameters from the static parameter database;

[0052] The dynamic elastic modulus is input into the static converter, and a linear transformation operation is performed by combining the dynamic-static conversion slope parameter and the intercept parameter to generate the static reference modulus.

[0053] When the tree age is less than the species maturity threshold, a continuously adjustable tree age decay factor is constructed based on the S-shaped growth curve.

[0054] The static reference modulus is attenuated by a continuously adjustable age attenuation factor, and the final static elastic modulus is output.

[0055] Secondly, embodiments of the present invention provide a wood mechanical property prediction system based on standing tree stress wave data, comprising:

[0056] The data acquisition module is used to obtain data on longitudinal wave velocity, green wood density, moisture content, forest germplasm, and tree age of living trees by performing physical measurements, biomarker analysis, and record association analysis on living trees.

[0057] The wave velocity correction module is used to screen for knot defects based on the abrupt response of the wave velocity gradient in the longitudinal wave velocity data of live trees, and to perform dynamic correction in combination with the moisture content of the live trees, and output standard longitudinal wave velocity data.

[0058] The wave velocity conversion module is used to perform boundary effect compensation on the standard longitudinal wave velocity data according to the boundary effect parameters determined by the tree germplasm, generate the log equivalent wave velocity, and perform gradient correction in combination with the radial spatial distribution relationship between the detection node and the pith, and output the corrected wave velocity parameters.

[0059] The dynamic modulus prediction module is used to perform multi-source coupling between the corrected wave velocity parameter and the raw material density, and the introduced density and wave velocity coupling coefficient, to construct a dynamic elastic modulus prediction model and predict the output dynamic elastic modulus.

[0060] The static modulus conversion module is used to convert the dynamic elastic modulus into a static reference modulus based on the growth characteristics corresponding to the tree germplasm, and to apply continuous attenuation correction by introducing a continuously adjustable tree age attenuation factor according to the tree age condition, so as to obtain the final static elastic modulus.

[0061] Thirdly, embodiments of the present invention provide a wood mechanical property prediction device based on standing tree stress wave data, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the wood mechanical property prediction method based on standing tree stress wave data as described above.

[0062] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the wood mechanical property prediction method based on standing timber stress wave data as described above.

[0063] (III) Beneficial Effects

[0064] The beneficial effects of this invention are: by using a systematic coupling and dynamic correction mechanism of multi-source stress wave data, this invention eliminates interference factors and significantly improves the accuracy and applicability of predicting the elastic modulus of living trees, achieving the goal of accurately assessing the material properties of living trees without logging.

[0065] First, by integrating physical measurements (wave velocity, density, moisture content), biomarker analysis (DNA / spectroscopy), and stand archive data, we can accurately and with low damage obtain multi-dimensional physical parameters of living trees, and reduce ecological disturbance while maintaining the integrity of the tree trunk structure, thus providing benchmark data for subsequent analysis.

[0066] Secondly, by screening for node defects through abrupt changes in wave velocity gradient response and combining this with dynamic moisture content correction to compensate for environmental interference, the signal distortion problem caused by differences in biological characteristics in single-parameter models is effectively overcome, thus improving the reliability of wave velocity characterization. Furthermore, based on boundary effect compensation and radial gradient correction using forest tree germplasm, the nonlinear influence of differences in wood structure on wave velocity propagation is further eliminated, enhancing the model's adaptability.

[0067] Furthermore, by introducing a density and wave velocity coupling coefficient to construct a dynamic prediction model, and simultaneously integrating the interaction between multiple source parameters, the systematic bias caused by the traditional method ignoring the coupling relationship of material properties is solved.

[0068] Finally, by combining the growth characteristics of forest tree germplasm to convert static baseline modulus and applying continuously adjustable attenuation correction factors based on tree age conditions, the dependence of traditional models on specific tree ages or single forest tree germplasm is overcome, and accurate adaptation of elastic modulus prediction across tree ages and multiple forest tree germplasms is achieved.

[0069] This invention combines versatility and specificity. It can cover a wide range of tree species scenarios through parameter range design, and optimize the prediction accuracy for key forest tree germplasm, providing scientific decision support for forest tree breeding and resource classification. Attached Figure Description

[0070] Figure 1 This is a schematic diagram of the overall process of the method provided in the embodiments of the present invention;

[0071] Figure 2 This is a schematic diagram illustrating the specific process of step S1 of the method provided in this embodiment of the invention;

[0072] Figure 3 This is a schematic diagram illustrating the operation of the north-south bidirectional measurement method provided in this embodiment of the invention;

[0073] Figure 4 This is a schematic diagram illustrating the specific process of step S2 in the method provided in this embodiment of the invention;

[0074] Figure 5 This is a schematic flowchart illustrating step S24 of the method provided in this embodiment of the invention.

[0075] Figure 6 This is a detailed flowchart illustrating step S3 of the method provided in this embodiment of the invention.

[0076] Figure 7 This is a detailed flowchart illustrating step S4 of the method provided in this embodiment of the invention;

[0077] Figure 8 A schematic diagram of the specific process of step S5 of the method provided in the embodiment of the present invention. Detailed Implementation

[0078] To better explain and facilitate understanding of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0079] like Figure 1As shown in the embodiment of the present invention, a method for predicting the mechanical properties of wood based on stress wave data of living trees includes: obtaining longitudinal wave velocity data, green wood density, moisture content, forest germplasm, and tree age of living trees through physical measurement, biomarker analysis, and archive association analysis; screening for node defects based on wave velocity gradient abrupt response of the longitudinal wave velocity data of living trees, and performing dynamic correction in combination with the moisture content of living trees to output standard longitudinal wave velocity data; performing boundary effect compensation on the standard longitudinal wave velocity data according to the boundary effect parameters determined by the forest germplasm to generate the equivalent wave velocity of logs, and performing gradient correction in combination with the radial spatial distribution relationship between the detection nodes and the pith to output corrected wave velocity parameters; performing multi-source coupling of the corrected wave velocity parameters with green wood density and the introduced density-wave velocity coupling coefficient to construct a dynamic elastic modulus prediction model and predict and output the dynamic elastic modulus; converting the dynamic elastic modulus into a static reference modulus based on the growth characteristics corresponding to the forest germplasm, and applying continuous attenuation correction by introducing a continuously adjustable tree age attenuation factor according to the tree age condition to obtain the final static elastic modulus.

[0080] This invention significantly improves the accuracy and applicability of predicting the elastic modulus of living timber through a systematic coupling and dynamic correction mechanism of multi-source stress wave data.

[0081] First, by integrating physical measurements (wave velocity, density, moisture content), biomarker analysis (DNA / spectroscopy), and stand archive data, we can accurately and with low damage obtain multi-dimensional physical parameters of living trees, and reduce ecological disturbance while maintaining the integrity of the tree trunk structure, thus providing benchmark data for subsequent analysis.

[0082] Secondly, by screening for node defects through abrupt changes in wave velocity gradient response and combining this with dynamic moisture content correction to compensate for environmental interference, the signal distortion problem caused by differences in biological characteristics in single-parameter models is effectively overcome, thus improving the reliability of wave velocity characterization. Furthermore, based on boundary effect compensation and radial gradient correction using forest tree germplasm, the nonlinear influence of differences in wood structure on wave velocity propagation is further eliminated, enhancing the model's adaptability.

[0083] Furthermore, by introducing a density and wave velocity coupling coefficient to construct a dynamic prediction model, and simultaneously integrating the interaction between multiple source parameters, the systematic bias caused by the traditional method ignoring the coupling relationship of material properties is solved.

[0084] Finally, by combining the growth characteristics of forest tree germplasm to convert static baseline modulus and applying continuously adjustable attenuation correction factors based on tree age conditions, the dependence of traditional models on specific tree ages or single forest tree germplasm is overcome, and accurate adaptation of elastic modulus prediction across tree ages and multiple forest tree germplasms is achieved.

[0085] This invention combines versatility and specificity. It can cover a wide range of tree species scenarios through parameter range design, and optimize the prediction accuracy for key forest tree germplasm, providing scientific decision support for forest tree breeding and resource classification.

[0086] To better understand the above technical solutions, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that the present invention can be understood more clearly and thoroughly, and that the scope of the present invention can be fully conveyed to those skilled in the art.

[0087] Specifically, embodiments of the present invention provide a method for predicting the mechanical properties of wood based on stress wave data of living trees, including:

[0088] S1. By conducting physical measurements, biomarker analysis, and record association analysis on living trees, we obtain data on longitudinal wave velocity, green wood density, moisture content, tree germplasm, and tree age of living trees.

[0089] Furthermore, such as Figure 2 As shown, step S1 includes:

[0090] S11. Control the mechanical guiding device to penetrate the bark to a predetermined depth in the sapwood layer at the preset detection node deployment position, thereby obtaining the standardized installation guide hole for the stress wave sensor. Control the mechanical guiding device (such as a hydraulic piercing mechanism) to penetrate the bark to a depth of 20.0-40.0 mm at the preset detection node in the diameter-at-breast section of the standing tree, thereby obtaining the standardized installation guide hole for the stress wave sensor.

[0091] S12. Based on the installation location of the stress wave sensor, guide the stress wave sensor to be deployed along the north-south axis of the tree trunk within the 0.5-2.0m height range of the standing tree's diameter at breast height (the zone of most stable wave velocity), completing the deployment. (Reference) Figure 3 It can be seen that stress wave sensor groups are deployed along the north-south axis of the tree trunk at a spacing of 1000±1mm. Specifically, they are microsecond stress wave detectors (accuracy ±1μs), ensuring that the first type of sensor is located in the due south longitudinal direction and the second type of sensor is located in the due north longitudinal direction, with a topological spacing calibration error of <0.1%.

[0092] S13. Control the standard impact hammer to perform vertical impact excitation, and simultaneously acquire the north-south bidirectional stress wave propagation time series data stream. Drive the standard impact hammer to perform vertical impact (once per second), with the hammer angle set at 45° and the deviation controlled at <2°. Simultaneously trigger the stress wave sensor to record timestamps and capture the north-south bidirectional stress wave propagation time series.

[0093] S14. Drive the sensor group to repeatedly acquire data streams N times. Obtain the effective propagation time by discarding abnormal data with fluctuation amplitudes exceeding 10%, and calculate the longitudinal wave velocity of the living tree by combining the topological spacing of the sensors. Each detection should repeat at least 3 complete impact-acquisition cycles.

[0094] S15. Control the drilling of wood core samples at the diameter at breast height (DBH) using a growth cone. Simultaneously collect the mass and three-dimensional volume data of the wood core samples to determine the green wood density. Drill wood cores at the DBH of the living standing trees using a growth cone, calculate the volume using the volume formula, and weigh the samples. Finally, calculate the density using ρ=M / V, where ρ is the green wood density (g / cm³). 3 M is the mass of the wood core, and V is the volume of the wood core.

[0095] S16. The moisture content is measured in real time by a pre-deployed moisture content sensor. After verifying the validity of the data, the current moisture content value is output. If a communication interruption or data drift is detected in the moisture content sensor, the corresponding moisture content in the preset green material moisture content parameter library is called as the current moisture content. Specifically, the validity of the data from the pre-deployed moisture content sensor is verified. When the confidence index of the moisture content sensor is continuously within the preset threshold range, the current moisture content is obtained by parallel measurement of the fusion resistance method and dielectric spectrum method. If a communication interruption or abnormal data drift is detected in the multi-modal moisture content sensor, the system automatically switches to the pre-stored green material moisture content benchmark parameter library and retrieves the corresponding default moisture content range value as the current moisture content. Generally, the default moisture content is 120-150%.

[0096] S17. Read the identification code pre-installed on the standing trees to analyze the tree germplasm. If the identification code is damaged, initiate real-time comparison of near-infrared spectral features and DNA marker database to determine the tree germplasm. Read the tree germplasm code stored in the RFID or QR code electronic tag pre-embedded in the rhizosphere of the standing trees, and obtain the tree germplasm identifier through the decoder (e.g., strain D corresponds to "Japanese larch × Changbai larch 77-3"). When physical damage or signal loss of the RFID or QR code tag is detected, automatically perform near-infrared spectral feature scanning, extract the wood chemical fingerprint spectrum and perform feature matching with the pre-stored DNA marker database, and determine the tree germplasm category through maximum similarity comparison.

[0097] S18. Call the forest stand file to match the spatial coordinates of the current living trees. If there is a historical record, directly extract the tree age; otherwise, generate a reverse growth ring count through annual ring analysis to determine the tree age.

[0098] S2. Based on the abrupt change in wave velocity gradient response, the longitudinal wave velocity data of the living trees is used to trigger the screening of knot defects, and dynamic correction is performed in combination with the moisture content of the living trees to output standard longitudinal wave velocity data.

[0099] Furthermore, such as Figure 4As shown, step S2 includes:

[0100] S21. Perform wavelet packet decomposition on the longitudinal wave velocity data of the living trees to obtain the multi-band wave velocity feature components corresponding to each detection node.

[0101] S22. Select the target component of the preset main frequency band in each node from the multi-band wave velocity feature components, and perform wavelet threshold denoising processing.

[0102] S23. Perform synchronous squeezing wavelet transform on the denoised target component signal to obtain the time spectrum aligned with adjacent detection nodes, detect the jump points on the time spectrum (spatial discontinuities caused by sensor deployment errors), and correct the jump points on the time spectrum by interpolation.

[0103] S24. Based on the corrected time-frequency spectrum, calculate the wave velocity gradient sequence according to the longitudinally equidistant topological structure of the tree trunk, and map it to a three-dimensional cylindrical coordinate system to generate a continuous wave velocity gradient matrix. The continuous wave velocity gradient matrix is ​​constructed based on the three-dimensional cylindrical coordinate system through longitudinal, circumferential, and radial discretization, and incorporates functions that integrate the propagation characteristics of stress waves in the wood fiber structure and the radial attenuation law. It is used to characterize the gradient field where wave velocity continuously changes with spatial location.

[0104] In one specific embodiment, the longitudinal wave velocity data of the living trees is decomposed into 8 layers of wavelet packets using the Daubechies-6 wavelet basis, and 256 wave velocity characteristic components in each frequency band are separated layer by layer to eliminate the frequency aliasing effect caused by sensor installation angle deviation (±3°) or contact impedance difference (<50Ω).

[0105] Next, the main vibration mode frequency band (0.5-10kHz) of each detection node was screened through energy entropy analysis, and high-frequency noise (>10kHz) and low-frequency artifacts (<0.5kHz) were removed. The resulting target components are: Where E represents the energy percentage of the frequency band. The main frequency band phase angle.

[0106] Dynamically select the target component of the main frequency band from the multi-frequency components. If the density of the forest tree species is higher than the predetermined density (such as larch strain D: Japanese larch × Changbai larch 77-3, where the density is directly regulated by the cellulose / lignin ratio determined by the strain genome), the 2-5kHz component is preferred. If the density of the forest tree species is not higher than the predetermined density (such as larch strain A: Japanese larch 3 × Dahurian larch 2), the 5-10kHz component is locked.

[0107] Next, wavelet thresholding denoising is performed using a piecewise threshold function:

[0108] ;

[0109] In the formula, λ is the robustness coefficient, which is an empirical value based on the statistical characteristics of noise standard deviation, λ=0.6745, and σ is the noise standard deviation, estimated by the signal extreme value statistical method, σ=median(|F i |) / λ;T h As a hard threshold, wavelet coefficients whose absolute value exceeds this value are considered valid signals and are directly retained (only the noise portion is reduced). T h =3σ,T s For soft threshold, T s =1.5σ, between T s With T h The wavelet coefficients between the given values ​​are subjected to soft-thresholding shrinkage. The "+" sign indicates that the positive part of the value within the parentheses is taken (i.e., if the result is negative, it is set to zero), ensuring that the wavelet coefficients after soft-thresholding are non-negative and avoiding sign flipping.

[0110] Next, a synchronous squeezing transformation is performed on the time spectrum of adjacent detection nodes (i, i+1), and the instantaneous frequency difference is calculated:

[0111] ;

[0112] When Δf(t) continuously deviates from the theoretical propagation template constructed by fitting historical data (deviation > ±5Hz), it is determined to be a sensor deployment error. A continuous phase field is reconstructed using cubic spline interpolation (node ​​spacing ≤ 20mm) or radial basis function interpolation (spacing > 20mm). Using the corrected time-spectrum data, polynomial interpolation (such as cubic polynomial interpolation) is used to restore the longitudinal continuity of the wave velocity gradient.

[0113] Furthermore, the tree trunk axis is divided at 5mm intervals, achieving an interpolation accuracy of ±0.1mm. Then, a cylindrical coordinate system is expanded with a resolution of Δθ=1°, extending from the surface of the log inwards, thus obtaining a continuous wave velocity gradient matrix:

[0114] ;

[0115] In the formula, z is the longitudinal coordinate along the trunk, with a resolution of Δz = 5 mm; θ is the circumferential angle, expanded into a cylindrical coordinate system with a resolution of Δθ = 1°; r is the radial depth, extending from the surface of the log inwards, with a resolution of Δr = 2 mm; k is the modal order, characterizing the harmonic order of the circumferential distribution, k = 1 corresponds to the fundamental frequency distribution, k = 2 corresponds to the second harmonic, etc.; N is the total number of modes, determined by the number of layers in the wavelet packet decomposition; V k The reference wave velocity value corresponding to the k-th mode is obtained by denoising and phase calibration of the wave velocity components of each frequency band extracted by wavelet packet decomposition; ω kαr is the modal weighting coefficient, dynamically allocated based on the energy proportion of each frequency band. α0 is the attenuation coefficient, characterizing the exponential attenuation of wave velocity with increasing radial depth r, with a value ranging from α0 = 0.03 (for tree species with a predetermined density) to α0 = 0.06 (for tree species with a predetermined density). It is important to understand that this formula is based on the premise that wood is a natural cylindrical structure and is expanded using cylindrical coordinates. Furthermore, considering that stress waves propagate within wood and are affected by fiber structure, density gradient, etc., the wave velocity typically exhibits an exponential decay characteristic along the radial direction. Therefore, an exponential decay term is introduced. .

[0116] S25. Real-time analysis of the longitudinal first derivative of the continuous wave velocity gradient matrix. When the absolute value of the derivative between adjacent nodes exceeds the adaptive threshold determined by the current timber density and tree germplasm, it is automatically marked as a knot defect region, and the longitudinal wave velocity data of the living trees corresponding to the knot defect region is removed.

[0117] The first derivative is calculated node-by-node along the longitudinal dimension of the wave velocity gradient matrix:

[0118] ;

[0119] Wherein, Δz is the longitudinal spacing, typically taken as 5 mm. When the absolute value of the derivative between adjacent detection nodes exceeds the adaptive threshold jointly determined by the current timber density and the inherent properties of the forest germplasm, it is identified as a node defect region, and the data of that node is automatically removed and the abnormal site is marked.

[0120] S26. Based on the deviation between the current moisture content and the reference moisture content, and in conjunction with the introduced moisture content correction coefficient, perform moisture content coupling compensation processing on the longitudinal wave velocity data of standing trees after removing knot defects, to obtain the primary corrected wave velocity. This step uses the following moisture content correction formula:

[0121] V dry =V live ×[1+η×(MC-MC0)];

[0122] Among them, V live Longitudinal wave velocity data of live standing trees to eliminate data with knot defects; MC is the measured moisture content of live standing trees (which can be obtained through multiple sources such as a resistance meter); The reference moisture content is taken as the median of 120%-150% in the green state by default; η is the moisture content correction coefficient, with a value range of 0.02 to 0.04.

[0123] S27. For the primary correction wave velocity, the anisotropic wave velocity distortion caused by the moisture absorption and expansion of wood fibers is compensated by nonlinear mapping, and standard longitudinal wave velocity data after moisture content effect stripping is generated.

[0124] Furthermore, such as Figure 5 As shown, step S24 specifically includes:

[0125] S271. Based on the longitudinal gradient distribution of the primary corrected wave velocity, the principal orientation angle of the fibers in living trees is extracted using continuous wavelet transform. By analyzing the longitudinal gradient distribution of the primary corrected wave velocity, the principal orientation angle of the fibers in living trees is extracted using continuous wavelet transform (using the Morlet wavelet basis). The wavelet transform captures the periodic fluctuation characteristics of the wave velocity, and the principal orientation angle of the fibers is determined by combining it with phase consistency analysis. The principal orientation angle of the fibers refers to the average tilt angle between the axial fiber arrangement direction of the living tree and the vertical line perpendicular to the longitudinal axis of the trunk.

[0126] S272. Based on the offset between the current moisture content and the reference moisture content, and the fiber principal orientation angle, dynamically construct a nonlinear compensation function that matches the rheological properties of the wood. Based on the rheological properties of the wood, and combining the current moisture content offset (ΔMC) and the fiber principal orientation angle, construct a dynamic nonlinear compensation function:

[0127] ;

[0128] In the formula, γ0 is the compensation coefficient determined by historical statistical data and is related to tree species density; β0 is the expansion coefficient determined by historical statistical data; ΔMC is the moisture content offset; and θ f The principal orientation angle of the fiber. The azimuth weighting factor is 1.2 for longitudinal direction and 0.8 for radial direction.

[0129] This function introduces a transverse hygroscopic expansion effect (proportional to ΔMC and the square of the fiber direction sine) and simulates strain saturation characteristics using a tanh function to match the nonlinear deformation behavior of wood. The compensation coefficient is dynamically adjusted according to tree species density to ensure differentiated correction between hardwood and softwood.

[0130] S273. Based on the north-south orientation of the stress wave detection nodes, apply differentiated azimuth weighting factors to the nonlinear compensation function; where the weight of the longitudinal propagation path is greater than that of the radial propagation path. Based on the north-south orientation of the stress wave detection nodes, differentiated weights are assigned to the compensation function: the weight of the longitudinal propagation path (north-south) is set to 1.2, and the weight of the radial path (east-west) is 0.8. This rule strengthens the correction of the longitudinal path (because fiber orientation has a more significant impact on longitudinal wave velocity) while suppressing the risk of overcorrection in the radial path.

[0131] S274. Input the primary corrected wave velocity into the nonlinear compensation function to perform anisotropic distortion correction, generating standard longitudinal wave velocity data after water content effect removal. Input the primary corrected wave velocity into the weighted nonlinear compensation function to perform water content expansion effect removal node by node. After correction, standard longitudinal wave velocity data is generated, and its anisotropic distortion rate is reduced.

[0132] S3. Perform boundary effect compensation on the standard longitudinal wave velocity data according to the boundary effect parameters determined by the tree germplasm to generate the log equivalent wave velocity, and perform gradient correction in combination with the radial spatial distribution relationship between the detection node and the pith to output the corrected wave velocity parameters.

[0133] Furthermore, such as Figure 6 As shown, step S3 includes:

[0134] S31. Perform boundary effect transformation processing on the standard longitudinal wave velocity data from standing trees to logs according to the boundary effect parameters determined by the tree germplasm, generating the log equivalent wave velocity. The boundary effect transformation formula is: V log =α×V std +β. In the formula, α is the wave velocity conversion coefficient from live larch standing trees to logs, with a value range of 1.12–1.25; β is the boundary effect compensation term, with a value range of -300 to -200 m / s, and the specific value is adjusted according to the forest tree species.

[0135] S32. Based on the bidirectional north-south stress wave propagation time series data stream, extract the stress wave propagation time difference of multiple detection nodes, and dynamically reconstruct the propagation path of the stress wave around the spinal cord based on the propagation time difference.

[0136] S33. Discretize the propagation path into multiple micro-segments with consistent fiber orientation, determine the main propagation direction (longitudinal / radial) of each micro-segment, and apply the corresponding orientation weighting factor. The propagation time of each micro-element is accumulated to obtain the theoretical propagation time of the stress wave to each node. By comparing it with the stress wave propagation time series data stream, a time difference residual matrix characterizing the spatial error distribution is generated.

[0137] S34. Based on the spatial error distribution characteristics of the time difference residual matrix, iteratively optimize the core coordinates until the preset convergence condition is met, and output the optimal core three-dimensional coordinates.

[0138] In one specific embodiment, the center of the largest inscribed circle of the trunk cross section is obtained by a laser rangefinder and used as the geometric center point of the trunk. The geometric center point of the trunk is regarded as the three-dimensional initial coordinates of the pith. Then, based on the three-dimensional initial coordinates of the pith, the propagation path of the stress wave from the impact point to each detection node is dynamically simulated. Specifically, the stress wave propagates along the main direction of the wood fiber around the pith in a curved (non-linear) manner, and the radius of curvature of the path is dynamically adjusted with the offset of the pith.

[0139] Then, the propagation path is discretized into multiple micro-segments with consistent fiber orientation. The main propagation direction (longitudinal / radial) of each micro-segment is determined, and a corresponding azimuth weighting factor is applied. Then, the propagation time of each micro-element is accumulated to obtain the theoretical propagation time of the stress wave to each node. By comparing it with the stress wave propagation time series data stream, a three-dimensional time difference residual matrix characterizing the spatial error distribution is generated. The smaller the residual value, the closer the assumed core position is to the actual position.

[0140] Furthermore, the core coordinates are iteratively optimized based on the spatial error distribution characteristics of the residual matrix until the preset convergence condition is met, and the optimal core three-dimensional coordinates are output.

[0141] Among them, for the assumed core coordinates (x c ,y c ,z c ), residual matrix elements R ijk A cubic mesh metric characterizing the difference between theoretically calculated propagation time and the measured time:

[0142] ;

[0143] In the formula, For the coordinates of discretized grid points in three-dimensional space, Let n be the theoretical arrival time of the nth node. Let N be the actual measured arrival time of the nth node, where N is the total number of nodes.

[0144] The time difference residual matrix maps the global time matching degree under different core hypothesis locations through spatial discretization, and its minimum point corresponds to the true core coordinates. The essence of iterative optimization is to search for the valley of the residual hypersurface, and the typical convergence condition is... (Gradient magnitude threshold). Represents the time difference residual matrix at three-dimensional grid points The gradient vector at that point.

[0145] S35. Analyze the radial spatial distance between the stress wave detection node and the core based on the optimal three-dimensional coordinates of the core, and determine the radial gradient correction factor based on the radial spatial distribution relationship. The radial gradient correction factor increases as the node position approaches the sapwood region and decreases as it approaches the heartwood region.

[0146] S36. Perform gradient compensation calculation on the equivalent wave velocity of the log based on the radial gradient correction factor, and output the corrected wave velocity parameter that combines boundary effect correction and radial gradient correction.

[0147] Considering that the sapwood wave velocity is higher than that of the heartwood (gradient difference of 5.7%), a radial gradient correction factor γ is introduced, thus obtaining the output corrected wave velocity parameter: V corrected =γ×V log(γ = 1.05-1.08, adjusted according to the distance of the test point from the pith) Wherein, γ is the radial weighting factor, with a value ranging from 0.95 to 1.08 for larch. The value in the near-heartwood region approaches the lower limit, and the value in the near-sapwood region approaches the upper limit. The specific value needs to be adjusted according to the distance of the test point from the pith.

[0148] S4. The corrected wave velocity parameter is coupled with the raw material density and the introduced density-wave velocity coupling coefficient through multiple sources to construct a dynamic elastic modulus prediction model and predict the output dynamic elastic modulus.

[0149] Furthermore, such as Figure 7 As shown, step S4 includes:

[0150] S41. Through regression analysis of multiple tree germplasm and tree age samples, the density-wave velocity coupling coefficient and elastic modulus compensation constant of each group are calculated in batches and stored in the dynamic parameter database.

[0151] In this step, living and unblemished specimens covering multiple tree species and ages are collected. Greenwood density and corrected wave velocity are measured in a standardized manner. Combined with measured dynamic elastic modulus data, regression analysis is performed grouped by tree species and age to calculate the density-wave velocity coupling coefficient for each group in batches. ) and the elastic modulus compensation constant corresponding to the residual ( For samples of unknown tree species or tree age, a general method based on dynamic fine-tuning of greenwood density is proposed. The values ​​are then stored in the database to support the elastic modulus prediction model.

[0152] S42. Based on the current tree species and age, match the corresponding density and wave velocity coupling coefficients and elastic modulus compensation constants in real time from the dynamic parameter database.

[0153] S43. Generate a primary mechanical response value based on the square product of the raw material density and the corrected wave velocity parameter. Multiply the primary mechanical response value by the density and wave velocity coupling coefficient, and then superimpose the elastic modulus compensation constant that is dynamically related to the physiological characteristics of the forest tree to output the predicted value of the dynamic elastic modulus.

[0154] S5. The dynamic elastic modulus is converted into a static reference modulus based on the growth characteristics corresponding to the tree germplasm, and a continuously adjustable age decay factor is introduced according to the age condition to apply continuous decay correction, so as to obtain the final static elastic modulus.

[0155] This invention utilizes the static elastic modulus, the most important mechanical property indicator obtained from the stress wave velocity of living trees. Based on this indicator value, the quality of timber from trees that have undergone cultivation measures can be assessed early and non-destructively without logging. The assessment results can provide a scientific basis for screening superior forest tree germplasm, grading and classifying timber, selecting suitable timber for appropriate use, and promoting efficient utilization.

[0156] Furthermore, such as Figure 8 As shown, step S5 includes:

[0157] S51. Through regression analysis of samples from multiple tree species and tree ages, the dynamic-to-static conversion slope parameters and intercept parameters for each group are calculated in batches and stored in the static parameter database. In this step, flawless samples from multiple tree species and tree ages are collected, and the dynamic and static elastic moduli are standardized and measured. A mapping relationship between dynamic and static elastic moduli is established according to tree species and tree age groups. Based on the dynamic modulus (independent variable) and static modulus (dependent variable) of each group of samples, linear regression is used to solve the dynamic-to-static conversion slope parameters and intercept parameters for each group in batches. Finally, the parameters are stored in the database to support the dynamic-to-static modulus conversion model.

[0158] S52. Based on the current tree species and tree age, extract the corresponding dynamic-static conversion slope parameters and intercept parameters from the static parameter database.

[0159] S53. Input the dynamic elastic modulus into the static converter, and perform a linear transformation operation by combining the dynamic-static conversion slope parameter and intercept parameter to generate the static reference modulus.

[0160] S54. When the tree age is less than the species maturity threshold, construct a continuously adjustable tree age decay factor based on the S-shaped growth curve.

[0161] S55. Apply attenuation correction to the static reference modulus based on a continuously adjustable tree age attenuation factor, and output the final static elastic modulus.

[0162] In another specific embodiment, the following physical quantity is considered as input: the corrected wave velocity parameter V corrected Tree age (T), tree species (S), and green timber density (ρ) green .

[0163] First, construct the dynamic elastic modulus prediction model:

[0164] ;

[0165] in, The density-wave velocity coupling coefficient ranges from 0.0010 to 0.0015. It is the elastic modulus compensation constant, with a value ranging from 0.2 to 0.5.

[0166] Next, perform static MOE conversion:

[0167] MOE static =a×MOE d +b;

[0168] Where a is the dynamic-to-static conversion slope parameter, with a value ranging from 0.80 to 0.90 for larch; b is the intercept parameter, with a value ranging from 2.0 to 3.5.

[0169] When the tree age is less than the threshold age, a tree age decay factor based on the Logistic function can be introduced, which has the following form: Where T0 is the threshold age, which is generally taken as 25, and τ is the decay rate constant, which controls the steepness of the decay process. The larger the value, the more rapid the decay. The value is taken as 0.2~0.5.

[0170] Generally, there is a positive correlation between the elastic modulus and bending strength of wood. In actual engineering or wood grading, regression formulas (such as MOR ≈ k0 × MOE, where k0 is an empirical coefficient) are often established through a large amount of experimental data or by using density as an intermediate parameter for inference.

[0171] In one specific embodiment, based on the above-described method for predicting the mechanical properties of wood based on stress wave data of living trees, a 32-year-old Larch tree of strain D was subjected to real-time field testing, and the data obtained are as follows:

[0172] First, the longitudinal wave velocity V, obtained and processed at a point 1.5m northward. live =4561m / s, green material density ρ green =0.610g / cm³.

[0173] Secondly, through the data processing workflow, the wave velocity data is first converted from standing trees to logs based on the boundary effect parameters of this strain (conversion coefficient α=1.18, compensation term β=-250m / s) to obtain the equivalent wave velocity V of the logs. log =1.18×4561-250=5137m / s.

[0174] Subsequently, based on the spatial distribution characteristics of the detection nodes, a radial gradient correction factor of 1.06 is applied to the sapwood region, and the corrected wave velocity V is output. corrected =1.06×5137=5445m / s (sidewood weight).

[0175] Furthermore, the corrected wave velocity and the raw material density were fused using a multi-source calculation with a coupling coefficient (0.0012) to obtain the dynamic elastic modulus MOE. d =0.0012×0.610×5445²+0.352=16.02GPa.

[0176] Finally, based on the growth characteristics of strain D, it was converted into the static reference elastic modulus MOE. static =0.847×16.02+2.641≈15.9GPa.

[0177] To verify the model's accuracy, the log was measured in the laboratory after felling. The results showed that the static elastic modulus was 16.02 GPa, with an error of less than 1% compared to the predicted value. This fully demonstrates the high accuracy and engineering applicability of this method in evaluating the mechanical properties of complex trees.

[0178] The embodiments of the present invention also provide some comparisons of material properties and parameter ranges.

[0179] Table 1 Comparison of Material Properties and Parameter Ranges

[0180]

[0181] Additionally, this invention provides a timber mechanical property prediction system based on standing tree stress wave data, comprising: a data acquisition module for acquiring standing tree longitudinal wave velocity data, green wood density, standing tree moisture content, forest germplasm, and tree age by performing physical measurements, biomarker analysis, and record association analysis on the standing trees; a wave velocity correction module for dynamically correcting the standing tree longitudinal wave velocity data based on wave velocity gradient abrupt response triggering node defect screening, and combining it with the standing tree moisture content to output standard longitudinal wave velocity data; and a wave velocity conversion module for converting the standard longitudinal wave velocity according to boundary effect parameters determined by the forest germplasm. The data performs boundary effect compensation to generate the equivalent wave velocity of the log, and performs gradient correction based on the radial spatial distribution relationship between the detection nodes and the pith, outputting the corrected wave velocity parameters. The dynamic modulus prediction module is used to couple the corrected wave velocity parameters with the green wood density and the introduced density-wave velocity coupling coefficient through multiple sources to construct a dynamic elastic modulus prediction model and predict the output dynamic elastic modulus. The static modulus conversion module is used to convert the dynamic elastic modulus into a static reference modulus based on the growth characteristics corresponding to the tree germplasm, and applies continuous attenuation correction by introducing a continuously adjustable age attenuation factor according to the tree age condition to obtain the final static elastic modulus.

[0182] Furthermore, embodiments of the present invention provide a timber mechanical property prediction device based on standing timber stress wave data, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the standing timber elastic modulus prediction method based on multi-source stress wave data as described above.

[0183] Then, this embodiment of the invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the wood mechanical property prediction method based on standing timber stress wave data as described above.

[0184] In summary, this invention provides a method, system, device, and medium for predicting the mechanical properties of wood based on stress wave data from living trees. The process is as follows: First, stress wave detection nodes are deployed in both north and south directions along the diameter at breast height (DBH) of living trees to collect longitudinal wave velocity data in situ without destructive processing. Simultaneously, multi-source parameters such as green wood density, moisture content, tree germplasm, and tree age are acquired. By performing abrupt change response analysis on the longitudinal wave velocity data, knot defect data is dynamically screened, and standard longitudinal wave velocity data is generated through dynamic correction based on moisture content. Subsequently, the standard wave velocity is compensated based on boundary effect parameters corresponding to the tree germplasm to generate equivalent wave velocities for logs. Further gradient correction is performed based on the radial spatial distribution relationship between the detection nodes and the pith, outputting corrected wave velocity parameters. On this basis, a dynamic elastic modulus prediction model is constructed. Finally, the dynamic elastic modulus is converted into a static reference modulus, and a continuously adjustable attenuation factor based on the Logistic function is introduced for correction, outputting a high-precision static elastic modulus.

[0185] This invention's non-destructive and efficient in-situ detection mechanism avoids damaging standing trees, protecting forest resources. Simultaneously, its cross-age prediction error is less than 8% (<12% for young timber), significantly better than the 15-20% error level of traditional single models. Furthermore, the forest germplasm-specific correction technology improves model adaptability by 30%, providing scientific support for the selection of high-quality structural timber and log grading (e.g., prioritizing 45-year-old timber with wave velocities >4500m / s for load-bearing structures). By combining a general framework with larch-specific parameters, a data-driven nonlinear logistic function age decay model is incorporated into the parameter range design. The robustness of the scheme is enhanced by a multi-source data fusion mechanism, achieving a complete innovation across the entire chain of standing tree mechanical performance evaluation, from theory to engineering application.

[0186] Since the systems / devices described in the above embodiments of the present invention are systems / devices used to implement the methods of the above embodiments of the present invention, those skilled in the art can understand the specific structure and modifications of the systems / devices based on the methods described in the above embodiments of the present invention, and therefore will not be repeated here. All systems / devices used in the methods of the above embodiments of the present invention fall within the scope of protection of the present invention.

[0187] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0188] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions.

[0189] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the claims should be interpreted to include both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0190] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, then this invention should also include these modifications and variations.

Claims

1. A method for predicting the mechanical properties of timber based on stress wave data of living trees, characterized in that, include: By conducting physical measurements, biomarker analysis, and record association analysis on living trees, we obtained data on longitudinal wave velocity, green wood density, moisture content, tree germplasm, and tree age of living trees. Based on the abrupt response of the wave velocity gradient, the longitudinal wave velocity data of the living trees is used to trigger the screening of knot defects, and dynamic correction is implemented in combination with the moisture content of the living trees to output standard longitudinal wave velocity data. Boundary effect compensation is performed on the standard longitudinal wave velocity data according to the boundary effect parameters determined by the tree germplasm to generate the log equivalent wave velocity. Gradient correction is then performed in combination with the radial spatial distribution relationship between the detection nodes and the pith to output the corrected wave velocity parameters. The dynamic elastic modulus prediction model is constructed by multi-source coupling of the corrected wave velocity parameter with the green wood density and the introduced density-wave velocity coupling coefficient, and the dynamic elastic modulus is predicted and output. This includes: calculating the density-wave velocity coupling coefficient and elastic modulus compensation constant for each group in batches through regression analysis of samples from multiple tree species and tree ages, and storing them in a dynamic parameter database; matching the corresponding density-wave velocity coupling coefficient and elastic modulus compensation constant in real time from the dynamic parameter database according to the current tree species and tree age; generating a primary mechanical response value based on the square product of the green wood density and the corrected wave velocity parameter, multiplying the primary mechanical response value by the current density-wave velocity coupling coefficient, and then superimposing the current elastic modulus compensation constant to predict and output the dynamic elastic modulus. The dynamic elastic modulus is converted into a static baseline modulus based on the growth characteristics corresponding to the tree germplasm, and a continuously adjustable age decay factor is introduced according to the age condition to apply continuous decay correction, thus obtaining the final static elastic modulus.

2. The method for predicting the mechanical properties of wood based on stress wave data of living trees as described in claim 1, characterized in that, Through physical measurements, biomarker analysis, and archival correlation analysis of living trees, data on longitudinal wave velocity, green wood density, moisture content, tree germplasm, and tree age were obtained, including: The mechanical guiding device is controlled to be deployed at a preset detection node position, penetrate the bark to reach a preset depth in the sapwood layer, and obtain the installation position of the stress wave sensor; Based on the installation location of the stress wave sensor, guide the stress wave sensor to be deployed in the section of the standing tree with a diameter at breast height of 0.5-2.0m, according to the topology of equal spacing along the north-south axis of the trunk. The standard impact hammer is controlled to perform vertical impact excitation, and the time series data stream of bidirectional stress wave propagation is acquired simultaneously from the north and south directions. The sensor group is driven to repeatedly collect data streams N times. The effective propagation time is obtained by removing abnormal data with fluctuation amplitude exceeding 10%. The longitudinal wave velocity data of the living tree is obtained by combining the topological spacing of the sensors. A growth cone was used to extract wood core samples from the diameter-to-breast-length section, and the mass and three-dimensional volume data of the wood core samples were collected simultaneously to determine the green wood density. The moisture content is measured in real time by a pre-deployed moisture content sensor. After verifying the validity of the data, the current moisture content value is output. If a communication interruption or data drift is detected in the moisture content sensor, the corresponding moisture content in the preset green material moisture content parameter library is called as the current moisture content. Read the identification code pre-set on the living tree to parse the forest tree germplasm. If the identification code is damaged, start the real-time comparison of near-infrared spectral features and DNA marker library to determine the forest tree germplasm. The system calls up the forest stand archive to match the spatial coordinates of the current living trees. If historical records exist, the tree age is extracted directly; otherwise, the tree age is determined by generating inverse growth ring counts through core ring analysis.

3. The method for predicting the mechanical properties of wood based on stress wave data of standing timber as described in claim 1, characterized in that, Longitudinal wave velocity data of living trees is used to trigger nodule defect screening based on abrupt changes in wave velocity gradient response, and dynamic correction is performed in conjunction with the moisture content of living trees. The output standard longitudinal wave velocity data includes: Wavelet packet decomposition was performed on the longitudinal wave velocity data of living trees to obtain the multi-band wave velocity feature components corresponding to each detection node; The target component of the preset main frequency band in each node is selected from the multi-band wave velocity feature components, and wavelet threshold denoising is performed. A synchronous squeezing wavelet transform is performed on the denoised target component signal to obtain the time spectrum aligned with adjacent detection nodes, the jump points on the time spectrum are detected, and the jump points on the time spectrum are corrected by interpolation. Based on the corrected time spectrum, the wave velocity gradient sequence is obtained according to the longitudinal equidistant topological structure of the tree trunk, and mapped to a three-dimensional cylindrical coordinate system to generate a continuous wave velocity gradient matrix. The continuous wave velocity gradient matrix is ​​constructed based on the three-dimensional cylindrical coordinate system through longitudinal, circumferential and radial discretization, and integrates the propagation characteristics of stress waves in the wood fiber structure and the radial attenuation law as functions, which are used to characterize the gradient field where the wave velocity changes continuously with spatial position. The longitudinal first derivative of the continuous wave velocity gradient matrix is ​​analyzed in real time. When the absolute value of the derivative between adjacent nodes exceeds the adaptive threshold determined by the current timber density and tree germplasm, it is automatically marked as a knot defect region, and the longitudinal wave velocity data of the living trees corresponding to the knot defect region is removed. Based on the deviation between the current moisture content and the reference moisture content, and combined with the introduced moisture content correction coefficient, the longitudinal wave velocity data of the standing trees after removing the knot defect data is subjected to moisture content coupling compensation processing to obtain the primary corrected wave velocity. For the primary correction wave velocity, standard longitudinal wave velocity data is generated by compensating for the anisotropic wave velocity distortion caused by the moisture absorption and expansion of wood fibers through nonlinear mapping.

4. The method for predicting the mechanical properties of wood based on stress wave data of living trees as described in claim 3, characterized in that, For the primary correction wave velocity, anisotropic wave velocity distortion caused by the hygroscopic expansion of wood fibers is compensated by nonlinear mapping to generate standard longitudinal wave velocity data, including: Based on the longitudinal gradient distribution of the primary correction wave velocity, the principal orientation angle of the living tree fiber is extracted by continuous wavelet transform; Based on the offset between the current moisture content and the reference moisture content and the fiber principal orientation angle, a nonlinear compensation function matching the rheological properties of wood is dynamically constructed. Based on the north-south orientation of the stress wave detection nodes, a differentiated orientation weighting factor is applied to the nonlinear compensation function; among which, the longitudinal propagation path weight is greater than the radial propagation path weight. The primary correction wave velocity is input into the nonlinear compensation function to perform anisotropic distortion correction, generating standard longitudinal wave velocity data.

5. The method for predicting the mechanical properties of timber based on stress wave data of living trees as described in claim 4, characterized in that, Boundary effect compensation is performed on the standard longitudinal wave velocity data according to the boundary effect parameters determined by the tree germplasm to generate the log equivalent wave velocity. Gradient correction is then performed based on the radial spatial distribution relationship between the detection nodes and the pith. The output corrected wave velocity parameters include: The standard longitudinal wave velocity data is converted from standing trees to logs using boundary effect parameters determined by tree germplasm to generate log equivalent wave velocities. Based on the acquired north-south bidirectional stress wave propagation time series data stream, the stress wave propagation time difference of multiple detection nodes is extracted, and the propagation path of the stress wave around the spinal cord is dynamically reconstructed based on the propagation time difference. The propagation path is discretized into multiple micro-segments with consistent fiber orientation. The main propagation direction of each micro-segment is determined, and a corresponding azimuth weighting factor is applied. The propagation time of each micro-segment is accumulated to obtain the theoretical propagation time of the stress wave to each node. By comparing it with the stress wave propagation time series data stream, a time difference residual matrix characterizing the spatial error distribution is generated. Based on the spatial error distribution characteristics of the time difference residual matrix, the core coordinates are iteratively optimized until the preset convergence condition is met, and the optimal core 3D coordinates are output. The radial spatial distance between the stress wave detection node and the core is analyzed based on the optimal three-dimensional coordinates of the core. The radial gradient correction factor is determined based on the radial spatial distribution relationship. The radial gradient correction factor increases as the node position approaches the sapwood region and decreases as it approaches the heartwood region. The equivalent wave velocity of the log is calculated based on the radial gradient correction factor, and the output corrected wave velocity parameter is obtained by fusing boundary effect correction and radial gradient correction.

6. The method for predicting the mechanical properties of wood based on stress wave data of living trees as described in claim 1, characterized in that, The dynamic elastic modulus is converted into a static baseline modulus based on the growth characteristics corresponding to the tree germplasm. A continuously adjustable age decay factor is then introduced to apply continuous decay correction based on tree age conditions, resulting in the final static elastic modulus, which includes: Through regression analysis of multiple tree germplasm and tree age samples, the dynamic-static transition slope parameters and intercept parameters of each group were calculated in batches and stored in the static parameter database. Based on the current tree species and tree age, extract the corresponding dynamic-static conversion slope parameters and intercept parameters from the static parameter database; The dynamic elastic modulus is input into the static converter, and a linear transformation operation is performed by combining the dynamic-static conversion slope parameter and the intercept parameter to generate the static reference modulus. When the tree age is less than the species maturity threshold, a continuously adjustable tree age decay factor is constructed based on the S-shaped growth curve. The static reference modulus is attenuated by a continuously adjustable age attenuation factor, and the final static elastic modulus is output.

7. A system for predicting the mechanical properties of timber based on stress wave data of standing trees, characterized in that, include: The data acquisition module is used to obtain data on longitudinal wave velocity, green wood density, moisture content, forest germplasm, and tree age of living trees by performing physical measurements, biomarker analysis, and record association analysis on living trees. The wave velocity correction module is used to screen for knot defects based on the abrupt response of the wave velocity gradient in the longitudinal wave velocity data of live trees, and to perform dynamic correction in combination with the moisture content of the live trees, and output standard longitudinal wave velocity data. The wave velocity conversion module is used to perform boundary effect compensation on the standard longitudinal wave velocity data according to the boundary effect parameters determined by the tree germplasm, generate the log equivalent wave velocity, and perform gradient correction in combination with the radial spatial distribution relationship between the detection node and the pith, and output the corrected wave velocity parameters. The dynamic modulus prediction module is used to perform multi-source coupling between the corrected wave velocity parameter and the green timber density, as well as the introduced density-wave velocity coupling coefficient, to construct a dynamic elastic modulus prediction model and predict the output dynamic elastic modulus. This includes: calculating the density-wave velocity coupling coefficient and elastic modulus compensation constant for each group in batches through regression analysis of samples from multiple tree species and ages, and storing them in the dynamic parameter database; matching the corresponding density-wave velocity coupling coefficient and elastic modulus compensation constant from the dynamic parameter database in real time based on the current tree species and age; generating a primary mechanical response value based on the square product of the green timber density and the corrected wave velocity parameter; multiplying the primary mechanical response value by the current density-wave velocity coupling coefficient; and then superimposing the current elastic modulus compensation constant to predict and output the dynamic elastic modulus. The static modulus conversion module is used to convert the dynamic elastic modulus into a static reference modulus based on the growth characteristics corresponding to the tree germplasm, and to apply continuous attenuation correction by introducing a continuously adjustable tree age attenuation factor according to the tree age condition, so as to obtain the final static elastic modulus.

8. A device for predicting the mechanical properties of wood based on stress wave data of standing timber, characterized in that, include: At least one processor; and memory that is communicatively connected to at least one processor; The memory stores instructions that can be executed by at least one processor, which enables the at least one processor to perform the wood mechanical property prediction method based on standing timber stress wave data as described in any one of claims 1-6.

9. A computer-readable storage medium storing computer-executable instructions thereon, characterized in that, When the executable instructions are executed by the processor, they implement the wood mechanical property prediction method based on stress wave data of living trees as described in any one of claims 1-6.

Images